1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.508 2007/05/14 18:56:25 magnus 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.
42 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
43 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
46 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
49 PH/02 In an ACL statement such as
51 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
53 if a client was not listed at all, or was listed with a value other than
54 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
55 the condition was not true (as it should be), so access was not denied.
56 The bug was that the ! inversion was incorrectly passed on to the second
57 item. This has been fixed.
59 PH/03 Added additional dnslists conditions == and =& which are different from
60 = and & when the dns lookup returns more than one IP address.
62 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
63 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
65 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
66 FSYNC, which compiles an option called disable_fsync that allows for
67 bypassing fsync(). The documentation is heavily laced with warnings.
69 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
71 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
72 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
73 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
74 including adding "make clean"; (3) Added -fPIC when compiling the test
75 dynamically loaded module, to get rid of a warning.
77 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
78 message fails, move_frozen_messages = true and ignore_bounce_errors_after
79 = 0s. The bug is otherwise harmless.
81 PH/07 There was a bug in the dovecot authenticator such that the value of
82 $auth1 could be overwritten, and so not correctly preserved, after a
83 successful authentication. This usually meant that the value preserved by
84 the server_setid option was incorrect.
86 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
88 PH/09 Installed PCRE release 7.0.
90 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
91 run for batched SMTP input. It is now run at the start of every message
92 in the batch. While fixing this I discovered that the process information
93 (output by running exiwhat) was not always getting set for -bs and -bS
94 input. This is fixed, and it now also says "batched" for BSMTP.
96 PH/11 Added control=no_pipelining.
98 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
99 patch, slightly modified), and move the expansion of helo_data till after
100 the connection is made in the smtp transport (so it can use these
103 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
105 PH/14 Added log_selector = +pid.
107 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
109 PH/16 Add ${if forany and ${if forall.
111 PH/17 Added dsn_from option to vary the From: line in DSNs.
113 PH/18 Flush SMTP output before performing a callout, unless control =
114 no_callout_flush is set.
116 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
117 was true (the default) a successful delivery failed to delete the retry
118 item, thus causing premature timeout of the address. The bug is now
121 PH/20 Added hosts_avoid_pipelining to the smtp transport.
123 PH/21 Long custom messages for fakedefer and fakereject are now split up
124 into multiline reponses in the same way that messages for "deny" and
125 other ACL rejections are.
127 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
128 with slight modification.
130 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
131 draft, changing the syntax and factoring some duplicate code".
133 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
134 for deliveries of the second and subsequent messages over the same SMTP
137 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
138 ${reduce, with only minor "tidies".
140 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
142 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
143 expansion side effects.
145 PH/27 When a message times out after an over-quota error from an Exim-imposed
146 quota, the bounce message says "mailbox is full". This message was not
147 being given when it was a system quota that was exceeded. It now should
150 MH/03 Made $recipients available in local_scan(). local_scan() already has
151 better access to the recipient list through recipients_list[], but
152 $recipients can be useful in postmaster-provided expansion strings.
154 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
155 in the case of a MAIL command with additional options following the
156 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
157 were accidentally chopped off.
159 PH/29 SMTP synchronization checks are implemented when a command is read -
160 there is a check that no more input is waiting when there shouldn't be
161 any. However, for some commands, a delay in an ACL can mean that it is
162 some time before the response is written. In this time, more input might
163 arrive, invalidly. So now there are extra checks after an ACL has run for
164 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
165 pipelining has not been advertised.
167 PH/30 MH's patch to allow iscntrl() characters to be list separators.
169 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
170 returned in the SMTP response when smtp_return_error_details was false.
173 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
174 instead of the C I/O that was originally supplied, because problems were
177 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
178 Exim which did not show up earlier: it was assuming that a call to
179 SSL_CTX_set_info_callback() might give an error value. In fact, there is
180 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
181 was a macro that became an assignment, so it seemed to work. This has
182 changed to a proper function call with a void return, hence the compile
183 error. Exim's code has been fixed.
185 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
188 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
190 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
192 PH/37 If a message is not accepted after it has had an id assigned (e.g.
193 because it turns out to be too big or there is a timeout) there is no
194 "Completed" line in the log. When some messages of this type were
195 selected by exigrep, they were listed as "not completed". Others were
196 picked up by some special patterns. I have improved the selection
197 criteria to be more general.
199 PH/38 The host_find_failed option in the manualroute router can now be set
200 to "ignore", to completely ignore a host whose IP address cannot be
201 found. If all hosts are ignored, the behaviour is controlled by the new
202 host_all_ignored option.
204 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
205 homing or because of multiple MX records with /mx) generated more than
206 one IP address, and the following item turned out to be the local host,
207 all the secondary addresses of the first item were incorrectly removed
208 from the list, along with the local host and any following hosts (which
209 is what is supposed to happen).
211 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
212 whoever called Exim into the -H file. In the case of the daemon it was
213 behaving confusingly. When first started, it used values for whoever
214 started the daemon, but after a SIGHUP it used the Exim user (because it
215 calls itself on a restart). I have changed the code so that it now always
218 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
219 message are rejected with the same error (e.g. no authentication or bad
220 sender address), and a DATA command is nevertheless sent (as can happen
221 with PIPELINING or a stupid MUA), the error message that was given to the
222 RCPT commands is included in the rejection of the DATA command. This is
223 intended to be helpful for MUAs that show only the final error to their
226 PH/42 Another patch from the Sieve maintainer.
228 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
229 Eximstats - Fixed some broken HTML links and added missing column headers
231 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
232 columns for Rejects, Temp Rejects, Ham, and Spam rows.
234 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
236 PH/43 Yet another patch from the Sieve maintainer.
238 PH/44 I found a way to check for a TCP/IP connection going away before sending
239 the response to the final '.' that terminates a message, but only in the
240 case where the client has not sent further data following the '.'
241 (unfortunately, this is allowed). However, in many cases there won't be
242 any further data because there won't be any more messages to send. A call
243 to select() can be used: if it shows that the input is "ready", there is
244 either input waiting, or the socket has been closed. An attempt to read
245 the next input character can distinguish the two cases. Previously, Exim
246 would have sent an OK response which the client would never have see.
247 This could lead to message repetition. This fix should cure that, at
248 least in a lot of common cases.
250 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
251 advertised in response to EHLO.
257 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
258 fixed by 4.65/MH/01 (is this a record?) are fixed:
260 (i) An empty string was always treated as zero by the numeric comparison
261 operators. This behaviour has been restored.
263 (ii) It is documented that the numeric comparison operators always treat
264 their arguments as decimal numbers. This was broken in that numbers
265 starting with 0 were being interpreted as octal.
267 While fixing these problems I realized that there was another issue that
268 hadn't been noticed. Values of message_size_limit (both the global option
269 and the transport option) were treated as octal if they started with 0.
270 The documentation was vague. These values are now always treated as
271 decimal, and I will make that clear in the documentation.
277 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
278 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
281 MH/01 Don't check that the operands of numeric comparison operators are
282 integers when their expansion is in "skipping" mode (fixes bug
283 introduced by 4.64-PH/07).
285 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
286 child addresses, Exim now panics and dies. Previously, because the count
287 is held in a short int, deliveries were likely to be lost. As such a
288 large number of recipients for a single message is ridiculous
289 (performance will be very, very poor), I have chosen to impose a limit
290 rather than extend the field.
296 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
297 leftover -K file (the existence of which was triggered by #402).
298 While we were at it, introduced process PID as part of the -K
299 filename. This should rule out race conditions when creating
302 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
303 processing considerably. Previous code took too long for large mails,
304 triggering a timeout which in turn triggers #401.
306 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
307 in the DK code in transports.c. sendfile() is not really portable,
308 hence the _LINUX specificness.
310 TF/01 In the add_headers option to the mail command in an Exim filter,
311 there was a bug that Exim would claim a syntax error in any
312 header after the first one which had an odd number of characters
315 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
316 callout verification, Exim cached a "reject" for the entire domain. This
317 is correct for most verifications, but it is not correct for a recipient
318 verification with use_sender or use_postmaster set, because in that case
319 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
320 case of MAIL FROM:<> rejection from other early rejections (e.g.
321 rejection of HELO). When verifying a recipient using a non-null MAIL
322 address, the cache is ignored if it shows MAIL FROM:<> rejection.
323 Whatever the result of the callout, the value of the domain cache is
324 left unchanged (for any other kind of callout, getting as far as trying
325 RCPT means that the domain itself is ok).
327 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
330 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
331 manifest itself as EPIPE rather than ECONNECT. When tidying away a
332 session, the daemon ignores ECONNECT errors and logs others; it now
333 ignores EPIPE as well.
335 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
336 (quoted-printable decoding).
338 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
339 later the small subsequent patch to fix an introduced bug.
341 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
343 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
345 PH/08 An error is now given if message_size_limit is specified negative.
347 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
348 to be given (somewhat) arbitrary names.
350 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
353 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
354 miscellaneous code fixes
356 PH/10 Added the log_reject_target ACL modifier to specify where to log
359 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
360 hostname. This is wrong, because it relates to the incoming message (and
361 probably the interface on which it is arriving) and not to the outgoing
362 callout (which could be using a different interface). This has been
363 changed to use the value of the helo_data option from the smtp transport
364 instead - this is what is used when a message is actually being sent. If
365 there is no remote transport (possible with a router that sets up host
366 addresses), $smtp_active_hostname is used.
368 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
369 tweaks were necessary in order to get it to work (see also 21 below):
370 (a) The code assumed that strncpy() returns a negative number on buffer
371 overflow, which isn't the case. Replaced with Exim's string_format()
373 (b) There were several signed/unsigned issues. I just did the minimum
374 hacking in of casts. There is scope for a larger refactoring.
375 (c) The code used strcasecmp() which is not a standard C function.
376 Replaced with Exim's strcmpic() function.
377 (d) The code set only $1; it now sets $auth1 as well.
378 (e) A simple test gave the error "authentication client didn't specify
379 service in request". It would seem that Dovecot has changed its
380 interface. Fortunately there's a specification; I followed it and
381 changed what the client sends and it appears to be working now.
383 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
386 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
387 address A is aliased to B and C, where B exists and C does not. Without
388 -v the output is "A verified" because verification stops after a
389 successful redirection if more than one address is generated. However,
390 with -v the child addresses are also verified. Exim was outputting "A
391 failed to verify" and then showing the successful verification for C,
392 with its parentage. It now outputs "B failed to verify", showing B's
393 parentage before showing the successful verification of C.
395 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
396 look up a TXT record in a specific list after matching in a combined
399 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
400 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
401 they consult the DNS. I had assumed they would set it the way they
402 wanted; and indeed my experiments on Linux seem to show that in some
403 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
404 To be on the safe side, however, I have now made the interface to
405 host_find_byname() similar to host_find_bydns(), with an argument
406 containing the DNS resolver options. The host_find_byname() function now
407 sets these options at its start, just as host_find_bydns() does. The smtp
408 transport options dns_qualify_single and dns_search_parents are passed to
409 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
410 of host_find_byname() use the default settings of RES_DEFNAMES
411 (qualify_single) but not RES_DNSRCH (search_parents).
413 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
414 spool_read_header() do less string testing, by means of a preliminary
415 switch on the second character of optional "-foo" lines. (This is
416 overdue, caused by the large number of possibilities that now exist.
417 Originally there were few.) While I was there, I also converted the
418 str(n)cmp tests so they don't re-test the leading "-" and the first
419 character, in the hope this might squeeze out yet more improvement.
421 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
422 flag allowing group syntax was set by the header_syntax check but not
423 turned off, possible causing trouble later; (2) The flag was not being
424 set at all for the header_verify test, causing "group"-style headers to
425 be rejected. I have now set it in this case, and also caused header_
426 verify to ignore an empty address taken from a group. While doing this, I
427 came across some other cases where the code for allowing group syntax
428 while scanning a header line wasn't quite right (mostly, not resetting
429 the flag correctly in the right place). These bugs could have caused
430 trouble for malformed header lines. I hope it is now all correct.
432 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
433 with the "reply" argument non-NULL. The code, however (which originally
434 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
435 but it didn't always do it. This confused somebody who was copying the
436 code for some other use. I have removed all the tests.
438 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
439 feature that was used to support insecure browsers during the U.S. crypto
440 embargo. It requires special client support, and Exim is probably the
441 only MTA that supported it -- and would never use it because real RSA is
442 always available. This code has been removed, because it had the bad
443 effect of slowing Exim down by computing (never used) parameters for the
444 RSA_EXPORT functionality.
446 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
447 authenticator to fail if there's a tab character in the incoming data
448 (there should never be unless someone is messing about, as it's supposed
449 to be base64-encoded). Also added, on Timo's advice, the "secured" option
450 if the connection is using TLS or if the remote IP is the same as the
451 local IP, and the "valid-client-cert option" if a client certificate has
454 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
455 authenticators. This can be used for authorization after authentication
456 succeeds. (In the case of plaintext, it servers for both authentication
459 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
460 if any retry times were supplied.
462 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
463 connection, where there is no client IP address to check. In this
464 situation, the verify now always succeeds.
466 PH/25 Applied John Jetmore's -Mset patch.
468 PH/26 Added -bem to be like -Mset, but loading a message from a file.
470 PH/27 In a string expansion for a processed (not raw) header when multiple
471 headers of the same name were present, leading whitespace was being
472 removed from all of them, but trailing whitespace was being removed only
473 from the last one. Now trailing whitespace is removed from each header
474 before concatenation. Completely empty headers in a concatenation (as
477 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
478 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
480 PH/29 [Removed. This was a change that I later backed out, and forgot to
481 correct the ChangeLog entry (that I had efficiently created) before
482 committing the later change.]
484 PH/30 Exim was sometimes attempting to deliver messages that had suffered
485 address errors (4xx response to RCPT) over the same connection as other
486 messages routed to the same hosts. Such deliveries are always "forced",
487 so retry times are not inspected. This resulted in far too many retries
488 for the affected addresses. The effect occurred only when there were more
489 hosts than the hosts_max_try setting in the smtp transport when it had
490 the 4xx errors. Those hosts that it had tried were not added to the list
491 of hosts for which the message was waiting, so if all were tried, there
492 was no problem. Two fixes have been applied:
494 (i) If there are any address or message errors in an SMTP delivery, none
495 of the hosts (tried or untried) are now added to the list of hosts
496 for which the message is waiting, so the message should not be a
497 candidate for sending over the same connection that was used for a
498 successful delivery of some other message. This seems entirely
499 reasonable: after all the message is NOT "waiting for some host".
500 This is so "obvious" that I'm not sure why it wasn't done
501 previously. Hope I haven't missed anything, but it can't do any
502 harm, as the worst effect is to miss an optimization.
504 (ii) If, despite (i), such a delivery is accidentally attempted, the
505 routing retry time is respected, so at least it doesn't keep
506 hammering the server.
508 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
509 in ${readsocket because some servers need this prod.
511 PH/32 Added some extra debug output when updating a wait-xxx database.
513 PH/33 The hint "could be header name not terminated by colon", which has been
514 given for certain expansion errors for a long time, was not being given
515 for the ${if def:h_colon_omitted{... case.
517 PH/34 The spec says: "With one important exception, whenever a domain list is
518 being scanned, $domain contains the subject domain." There was at least
519 one case where this was not true.
521 PH/35 The error "getsockname() failed: connection reset by peer" was being
522 written to the panic log as well as the main log, but it isn't really
523 panic-worthy as it just means the connection died rather early on. I have
524 removed the panic log writing for the ECONNRESET error when getsockname()
527 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
528 runs only) independently of the message's sender address. This meant
529 that, if the 4xx error was in fact related to the sender, a different
530 message to the same recipient with a different sender could confuse
531 things. In particualar, this can happen when sending to a greylisting
532 server, but other circumstances could also provoke similar problems.
533 I have changed the default so that the retry time for these errors is now
534 based a combination of the sender and recipient addresses. This change
535 can be overridden by setting address_retry_include_sender=false in the
538 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
539 remote server are returned as part of bounce messages. This was not
540 happening for LMTP over a pipe (the lmtp transport), but now it is the
541 same for both kinds of LMTP.
543 PH/38 Despite being documented as not happening, Exim was rewriting addresses
544 in header lines that were in fact CNAMEs. This is no longer the case.
546 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
547 and queue runs started by the daemon processed all messages. This has
548 been fixed so that -R and -S can now usefully be given with -q<time>.
550 PH/40 Import PCRE release 6.7 (fixes some bugs).
552 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
554 PH/42 Give an error if -q is specified more than once.
556 PH/43 Renamed the variables $interface_address and $interface_port as
557 $received_ip_address and $received_port, to make it clear that these
558 values apply to message reception, and not to the outgoing interface when
559 a message is delivered. (The old names remain recognized, of course.)
561 PH/44 There was no timeout on the connect() call when using a Unix domain
562 socket in the ${readsocket expansion. There now is.
564 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
565 be meaningful with "accept".
567 SC/01 Eximstats V1.43
568 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
570 SC/02 Eximstats V1.44
571 Use a glob alias rather than an array ref in the generated
572 parser. This improves both readability and performance.
574 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
575 Collect SpamAssassin and rejection statistics.
576 Don't display local sender or destination tables unless
577 there is data to show.
578 Added average volumes into the top table text output.
580 SC/04 Eximstats V1.46
581 Collect data on the number of addresses (recipients)
582 as well as the number of messages.
584 SC/05 Eximstats V1.47
585 Added 'Message too big' to the list of mail rejection
586 reasons (thanks to Marco Gaiarin).
588 SC/06 Eximstats V1.48
589 Mainlog lines which have GMT offsets and are too short to
590 have a flag are now skipped.
592 SC/07 Eximstats V1.49 (Alain Williams)
593 Added the -emptyok flag.
595 SC/08 Eximstats V1.50
596 Fixes for obtaining the IP address from reject messages.
598 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
599 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
600 whitesspace changes from 4.64-PH/27
602 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
605 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
606 are found, allow negative numbers in numeric criteria)
608 JJ/06 exipick.20061117.2, added new $message_body_missing variable
610 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
611 to match changes made in 4.64-PH/43
613 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
615 PH/47 Put in an explicit test for a DNS lookup of an address record where the
616 "domain" is actually an IP address, and force a failure. This locks out
617 those revolvers/nameservers that support "A-for-A" lookups, in
618 contravention of the specifications.
620 PH/48 When a host name was looked up from an IP address, and the subsequent
621 forward lookup of the name timed out, the host name was left in
622 $sender_host_name, contrary to the specification.
624 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
625 restricted to single-key lookups, Exim was not diagnosing an error if
626 * or *@ was used with a query-style lookup.
628 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
630 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
631 long ago, but noone interested enough thought of it. Let's just say that
632 the "1.1" means that there are some new functions that weren't there at
633 some point in the past.
635 PH/51 Error processing for expansion failure of helo_data from an smtp
636 transport during callout processing was broken.
638 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
639 tested/used via the -bh/-bhc/-bs options.
641 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
642 bug, fixed in subsequent PCRE releases).
644 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
645 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
647 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
653 SC/01 Use a glob alias rather than an array ref in eximstats generated
654 parser. This improves both readability and performance.
656 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
657 Don't display local sender or destination tables in eximstats unless
658 there is data to show.
659 Added average volumes into the eximstats top table text output.
661 SC/03 Collect data on the number of addresses (recipients) as well
662 as the number of messages in eximstats.
664 TF/01 Correct an error in the documentation for the redirect router. Exim
665 does (usually) call initgroups() when daemonizing.
667 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
668 with consistent privilege compared to when running as a daemon.
670 TF/03 Note in the spec that $authenticated_id is not set for local
671 submissions from trusted users.
673 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
674 Thanks to Dean Brooks <dean@iglou.com> for the patch.
676 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
677 by adding some example configuration directives to the default
678 configuration file. A little bit of work is required to uncomment the
679 directives and define how usernames and passwords are checked, but
680 there is now a framework to start from.
682 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
683 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
684 without this. I don't know how relevant this is to other LDAP libraries.
686 PH/02 Add the verb name to the "unknown ACL verb" error.
688 PH/03 Magnus Holmgren's patch for filter_prepend_home.
690 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
692 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
693 directory not expanded when it should be if an expanded home directory
694 was set for the address (which is overridden by the transport).
696 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
699 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
700 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
701 because it is too late at that time, and has no effect.
703 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
704 security issue with \' (bugzilla #107). I could not use the
705 PQescapeStringConn() function, because it needs a PGconn value as one of
708 PH/08 When testing addresses using -bt, indicate those final addresses that
709 are duplicates that would not cause an additional delivery. At least one
710 person was confused, thinking that -bt output corresponded to deliveries.
711 (Suppressing duplicates isn't a good idea as you lose the information
712 about possibly different redirections that led to the duplicates.)
714 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
715 systems where poll() doesn't work, in particular OS X.
717 PH/10 Added more information to debugging output for retry time not reached.
719 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
720 operations in malware.c.
722 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
725 PH/13 If write_rejectlog was set false when logging was sent to syslog with
726 syslog_duplication set false, log lines that would normally be written
727 both the the main log and to the reject log were not written to syslog at
730 PH/14 In the default configuration, change the use of "message" in ACL warn
731 statements to "add_header".
733 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
734 not followed by a command (e.g. "seen endif").
736 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
737 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
740 PH/17 Added extra conditions to the default value of delay_warning_condition
744 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
745 { match{$h_precedence:}{(?i)bulk|list|junk} } \
746 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
749 The Auto-Submitted: and various List- headers are standardised, whereas I
750 don't think Precedence: ever was.
752 PH/18 Refactored debugging code in route_finduser() to show more information,
753 in particular, the error code if getpwnam() issues one.
755 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
756 This is apparently needed in addition to the PH/07 change above to avoid
757 any possible encoding problems.
759 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
760 but not after initializing Perl.
762 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
763 output them only if debugging. By default they are written stderr,
764 apparently, which is not desirable.
766 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
769 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
772 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
774 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
775 authenticated or an ident call has been made. Suppress the default
776 values for $authenticated_id and $authenticated_sender (but permit -oMai
777 and -oMas) when testing with -bh.
779 PH/24 Re-jigged the order of the tests in the default configuration so that the
780 tests for valid domains and recipients precede the DNS black list and CSA
781 tests, on the grounds that those ones are more expensive.
783 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
784 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
785 This bug exists in every version of Exim that I still have, right back to
788 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
789 However, an attempt to turn on case-sensitivity in a regex key by
790 including (?-i) didn't work because the subject string was already
791 lowercased, and the effects were non-intuitive. It turns out that a
792 one-line patch can be used to allow (?-i) to work as expected.
798 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
799 other effects) broke the use of negated acl sub-conditions.
801 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
804 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
805 "Deny" causes Exim to reject the incoming connection with a 554 error.
806 Unfortunately, if there is a major crisis, such as a disk failure,
807 tcp-wrappers gives "deny", whereas what one would like would be some
808 kind of temporary error. A kludge has been added to help with this.
809 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
810 554 error is used if errno is still zero or contains ENOENT (which occurs
811 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
814 PH/03 Add -lutil to the default FreeBSD LIBS setting.
816 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
817 errors. Otherwise a message that provokes a temporary error (when other
818 messages do not) can cause a whole host to time out.
820 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
821 addresses were routed directly to files or pipes from a redirect router.
822 File deliveries just didn't batch; pipe deliveries might have suffered
825 PH/06 A failure to get a lock for a hints database would erroneously always say
826 "Failed to get write lock", even when it was really a read lock.
828 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
829 of 0600. This has been changed to use the value of the lockfile_mode
830 option (which defaults to 0600).
832 PH/08 Applied small patch from the Sieve maintainer.
834 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
835 folder from quota calculations, a direct delivery into this folder messed
836 up the contents of the maildirsize file. This was because the regex was
837 used only to exclude .Trash (or whatever) when the size of the mailbox
838 was calculated. There was no check that a delivery was happening into an
839 excluded directory. This bug has been fixed by ignoring all quota
840 processing for deliveries into excluded directories.
842 PH/10 Added the maildirfolder_create_regex option to appendfile.
848 PH/01 The code for finding all the local interface addresses on a FreeBSD
849 system running IPv6 was broken. This may well have applied to all BSD
850 systems, as well as to others that have similar system calls. The broken
851 code found IPv4 interfaces correctly, but gave incorrect values for the
852 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
853 that it would not match correctly against @[] and not recognize the IPv6
856 PH/02 The ipliteral router was not recognizing addresses of the form user@
857 [ipv6:....] because it didn't know about the "ipv6:" prefix.
859 PH/03 Added disable_ipv6.
861 PH/04 Changed $reply_address to use the raw form of the headers instead of the
862 decoded form, because it is most often used to construct To: headers
863 lines in autoreplies, and the decoded form may well be syntactically
864 invalid. However, $reply_address has leading white space removed, and all
865 newlines turned into spaces so that the autoreply transport does not
868 PH/05 If group was specified without a user on a router, and no group or user
869 was specified on a transport, the group from the router was ignored.
871 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
872 for visible compile-time settings that can be used to change these
873 numbers, for those that want even more. Backwards compatibility with old
874 spool files has been maintained. However, going back to a previous Exim
875 release will lost any variables that are in spool files.
877 PH/07 Two small changes when running in the test harness: increase delay when
878 passing a TCP/IP connection to a new process, in case the original
879 process has to generate a bounce, and remove special handling of
880 127.0.0.2 (sic), which is no longer necessary.
882 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
883 be the same on different OS.
885 PH/09 Moved a debug statement in filter processing to avoid a race problem when
888 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
889 whether --show-vars was specified or not
891 JJ/02 exipick: Added support for new ACL variable spool format introduced
894 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
895 syntactically invalid From: or Reply-to: line, and a filter used this to
896 generate an autoreply, and therefore failed to obtain an address for the
897 autoreply, Exim could try to deliver to a non-existent relative file
898 name, causing unrelated and misleading errors. What now happens is that
899 it logs this as a hard delivery error, but does not attempt to create a
902 PH/11 The exinext utility has a -C option for testing purposes, but although
903 the given file was scanned by exinext itself; it wasn't being passed on
904 when Exim was called.
906 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
907 an end-of-file indication when reading a command response.
909 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
910 compiled. In many other places in Exim, IPv6 addresses are always
911 recognized, so I have changed this. It also means that IPv4 domain
912 literals of the form [IPV4:n.n.n.n] are now always recognized.
914 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
915 used if the router is not running as root, for example, when verifying at
916 ACL time, or when using -bh. The debugging output from this situation was
917 non-existent - all you got was a failure to exec. I have made two
920 (a) Failures to set uid/gid, the current directory, or a process leader
921 in a subprocess such as that created by queryprogram now generate
922 suitable debugging ouput when -d is set.
924 (b) The queryprogram router detects when it is not running as root,
925 outputs suitable debugging information if -d is set, and then runs
926 the subprocess without attempting to change uid/gid.
928 PH/15 Minor change to Makefile for building test_host (undocumented testing
931 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
932 additional section of a DNS packet that returns MX or SRV records.
933 Instead, it always explicitly searches for A/AAAA records. This avoids
934 major problems that occur when a DNS server includes only records of one
935 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
936 fixed another bug: if SRV records were looked up and the corresponding
937 address records were *not* found in the additional section, the port
938 values from the SRV records were lost.
940 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
941 using the correct key (the original address) when searching the retry
942 rules in order to find which one to use for generating the retry hint.
944 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
945 adding the default one. Similarly, if it contains a Reply-To: header, the
946 errors_reply_to option, if set, is not used.
948 PH/19 When calculating a retry time, Exim used to measure the "time since
949 failure" by looking at the "first failed" field in the retry record. Now
950 it does not use this if it is later than than the arrival time of the
951 message. Instead it uses the arrival time. This makes for better
952 behaviour in cases where some deliveries succeed, thus re-setting the
953 "first failed" field. An example is a quota failure for a huge message
954 when small messages continue to be delivered. Without this change, the
955 "time since failure" will always be short, possible causing more frequent
956 delivery attempts for the huge message than are intended.
957 [Note: This change was subsequently modified - see PH/04 for 4.62.]
959 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
960 $1, $2, $3) because the numerical variables can be reset during some
961 expansion items (e.g. "match"), thereby losing the authentication data.
963 PH/21 Make -bV show the size of off_t variables so that the test suite can
964 decide whether to run tests for quotas > 2G.
966 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
967 mailbox_size, and mailbox_filecount in the appendfile transport. If a
968 filecount value is greater than 2G or if a quota value is greater than 2G
969 on a system where the size of off_t is not greater than 4, a panic error
972 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
973 never match. The debug and -bh output now contains an explicit error
974 message indicating a malformed IPv4 address or mask.
976 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
977 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
980 PH/25 Do not write to syslog when running in the test harness. The only
981 occasion when this arises is a failure to open the main or panic logs
982 (for which there is an explicit test).
984 PH/26 Added the /no_tell option to "control=freeze".
986 PH/27 If a host name lookup failed very early in a connection, for example, if
987 the IP address matched host_lookup and the reverse lookup yielded a name
988 that did not have a forward lookup, an error message of the form "no IP
989 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
990 could be logged. Now it outputs the IP address instead of "NULL".
992 PH/28 An enabling patch from MH: add new function child_open_exim2() which
993 allows the sender and the authenticated sender to be set when
994 submitting a message from within Exim. Since child_open_exim() is
995 documented for local_scan(), the new function should be too.
997 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
998 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
999 results in an empty string is now treated as unset.
1001 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1003 PH/31 Added sender_verify_fail logging option.
1005 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1006 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1007 tidied the source and removed it altogether.
1009 PH/33 When a queue run was abandoned because the load average was too high, a
1010 log line was always written; now it is written only if the queue_run log
1011 selector is set. In addition, the log line for abandonment now contains
1012 information about the queue run such as the pid. This is always present
1013 in "start" and "stop" lines but was omitted from the "abandon" line.
1015 PH/34 Omit spaces between a header name and the colon in the error message that
1016 is given when verify = headers_syntax fails (if there are lots of them,
1017 the message gets confusing).
1019 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1020 names, as there are now some PTR records that contain slashes. This check
1021 is only to protect against broken name servers that fall over on strange
1022 characters, so the fact that it applies to all lookups doesn't matter.
1024 PH/36 Now that the new test suite is complete, we can remove some of the
1025 special code in Exim that was needed for the old test suite. For example,
1026 sorting DNS records because real resolvers return them in an arbitrary
1027 order. The new test suite's fake resolver always returns records in the
1030 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1031 bounces) except when queue_only is set, to avoid logging races between
1032 the different processes.
1034 PH/38 Panic-die if .include specifies a non-absolute path.
1036 PH/39 A tweak to the "H" retry rule from its user.
1038 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1039 a label. They prevented compilation on older perls.
1041 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1042 a warning to be raised on newish perls.
1044 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1045 on queue. Changes to match documented behaviour of showing count of
1046 messages matching specified criteria.
1048 PH/40 Changed the default ident timeout from 30s to 5s.
1050 PH/41 Added support for the use of login_cap features, on those BSD systems
1051 that have them, for controlling the resources used by pipe deliveries.
1053 PH/42 The content-scanning code uses fopen() to create files in which to put
1054 message data. Previously it was not paying any attention to the mode of
1055 the files. Exim runs with umask(0) because the rest of the code creates
1056 files with open(), and sets the required mode explicitly. Thus, these
1057 files were ending up world-writeable. This was not a big issue, because,
1058 being within the spool directory, they were not world-accessible. I have
1059 created a function called modefopen, which takes an additional mode
1060 argument. It sets umask(777), creates the file, chmods it to the required
1061 mode, then resets the umask. All the relevant calls to fopen() in the
1062 content scanning code have been changed to use this function.
1064 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1065 to 24 hours. This avoids potential overflow problems when processing G
1066 and H retry rules. I suspect nobody ever tinkers with this value.
1068 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1070 PH/45 When the plaintext authenticator is running as a client, the server's
1071 challenges are checked to ensure they are valid base64 strings. By
1072 default, the authentication attempt is cancelled if an invalid string is
1073 received. Setting client_ignore_invalid_base64 true ignores these errors.
1074 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1075 they are received. Thus, the responses can be made to depend on the
1076 challenges. If an invalid string is ignored, an empty string is placed in
1079 PH/46 Messages that are created by the autoreply transport now contains a
1080 References: header, in accordance with RFCs 2822 and 3834.
1082 PH/47 Added authenticated_sender_force to the smtp transport.
1084 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1086 PH/49 Installed latest patch from the Sieve maintainer.
1088 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1089 was also set, the appendfile transport was unnecessarily scanning a
1090 directory of message files (e.g. for maildir delivery) to find the count
1091 of files (along with the size), even though it did not need this
1092 information. It now does the scan only if it needs to find either the
1093 size of the count of files.
1095 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1097 PH/52 Two bugs concerned with error handling when the smtp transport is
1100 (i) Exim was not creating retry information for temporary errors given
1101 for individual recipients after the DATA command when the smtp transport
1102 was used in LMTP mode. This meant that they could be retried too
1103 frequently, and not timed out correctly.
1105 (ii) Exim was setting the flag that allows error details to be returned
1106 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1107 recipients that were returned after the DATA command.
1109 PH/53 This is related to PH/52, but is more general: for any failing address,
1110 when detailed error information was permitted to be returned to the
1111 sender, but the error was temporary, then after the final timeout, only
1112 "retry timeout exceeded" was returned. Now it returns the full error as
1113 well as "retry timeout exceeded".
1115 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1116 do this, and (what is worse) MTAs that accept it.
1118 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1119 will now be deprecated.
1121 PH/56 New os.c-cygwin from the Cygwin maintainer.
1123 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1124 formats (previously only available in exim formats via -bpr, -bpru,
1125 and -bpra. Now also available in native and exiqgrep formats)
1127 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1128 with very large, slow to parse queues
1130 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1132 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1134 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1135 responses to authentication challenges, though it was showing the
1136 challenges; (ii) I've removed the CR characters from the debug output for
1139 PH/58 Allow for the insertion of a newline as well as a space when a string
1140 is turned into more than one encoded-word during RFC 2047 encoding. The
1141 Sieve code now uses this.
1143 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1144 data_4xx, lost_connection, tls_required.
1146 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1147 message was being sent as an SMTP response.
1149 PH/61 Add -l and -k options to exicyclog.
1151 PH/62 When verifying, if an address was redirected to one new address, so that
1152 verification continued, and the new address failed or deferred after
1153 having set something in $address_data, the value of $address_data was not
1154 passed back to the ACL. This was different to the case when no
1155 redirection occurred. The value is now passed back in both cases.
1157 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1158 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1159 use login_cap.h, so on its own it isn't the distinguishing feature. The
1160 new name refers directly to the setclassresources() function.
1162 PH/65 Added configuration files for NetBSD3.
1164 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1166 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1167 is preferred over IPv4.
1169 PH/68 The bounce_return_message and bounce_return_body options were not being
1170 honoured for bounces generated during the reception of non-SMTP messages.
1171 In particular, this applied to messages rejected by the ACL. This bug has
1172 been fixed. However, if bounce_return_message is true and bounce_return_
1173 body is false, the headers that are returned for a non-SMTP message
1174 include only those that have been read before the error was detected.
1175 (In the case of an ACL rejection, they have all been read.)
1177 PH/69 The HTML version of the specification is now built in a directory called
1178 spec_html instead of spec.html, because the latter looks like a path with
1179 a MIME-type, and this confuses some software.
1181 PH/70 Catch two compiler warnings in sieve.c.
1183 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1184 function verify_get_ident() calls ip_connect() to connect a socket, but
1185 if the "connect()" function timed out, ip_connect() used to close the
1186 socket. However, verify_get_ident() also closes the socket later, and in
1187 between Exim writes to the log, which may get opened at this point. When
1188 the socket was closed in ip_connect(), the log could get the same file
1189 descriptor number as the socket. This naturally causes chaos. The fix is
1190 not to close the socket in ip_connect(); the socket should be closed by
1191 the function that creates it. There was only one place in the code where
1192 this was missing, in the iplookup router, which I don't think anybody now
1193 uses, but I've fixed it anyway.
1195 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1196 well as to direct DNS lookups. Otherwise the handling of names in host
1197 lists is inconsistent and therefore confusing.
1203 PH/01 Two changes to the default runtime configuration:
1205 (1) Move the checks for relay_from_hosts and authenticated clients from
1206 after to before the (commented out) DNS black list checks.
1208 (2) Add control=submission to the relay_from_hosts and authenticated
1209 clients checks, on the grounds that messages accepted by these
1210 statements are most likely to be submissions.
1212 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1214 (1) Generate an error if the third argument for the ${prvs expansion is
1217 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1220 (3) Reset the variables that are obtained from the first argument of
1221 ${prvscheck and used in the second argument before leaving the code,
1222 because their memory is reclaimed, so using them afterwards may do
1225 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1226 one (it's much easier than Tom thought :-).
1228 (5) Because of (4), we can now allow for the use of $prvscheck_result
1229 inside the third argument.
1231 PH/03 For some reason, the default setting of PATH when running a command from
1232 a pipe transport was just "/usr/bin". I have changed it to
1235 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1236 anything to be listed in the output from -bV.
1238 PH/05 When a filter generated an autoreply, the entire To: header line was
1239 quoted in the delivery log line, like this:
1241 => >A.N.Other <ano@some.domain> <original@ddress> ...
1243 This has been changed so that it extracts the operative address. There
1244 may be more than one such address. If so, they are comma-separated, like
1247 => >ano@some.domain,ona@other.domain <original@ddress> ...
1249 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1250 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1251 not being looked up in the rDNS to get a host name, Exim was showing the
1252 IP address twice in Received: lines, even though the IP addresses were
1253 identical. For example:
1255 Received: from [1.2.3.4] (helo=[1.2.3.4])
1257 However, if the real host name was known, it was omitting the HELO data
1258 if it matched the actual IP address. This has been tidied up so that it
1259 doesn't show the same IP address twice.
1261 PH/07 When both +timestamp and +memory debugging was on, the value given by
1262 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1263 called by the debug printing, thereby overwriting the timestamp buffer.
1264 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1266 PH/08 When the original message was included in an autoreply transport, it
1267 always said "this is a copy of the message, including all the headers",
1268 even if body_only or headers_only was set. It now gives an appropriate
1271 PH/09 Applied a patch from the Sieve maintainer which:
1273 o fixes some comments
1274 o adds the (disabled) notify extension core
1275 o adds some debug output for the result of if/elsif tests
1276 o points to the current vacation draft in the documentation
1277 and documents the missing references header update
1281 o fixes a bug in processing the envelope test (when testing
1282 multiple envelope elements, the last element determinted the
1285 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1286 Electronic Mail") by including:
1288 Auto-submitted: auto-generated
1290 in the messages that it generates (bounce messages and others, such as
1291 warnings). In the case of bounce messages for non-SMTP mesages, there was
1292 also a typo: it was using "Auto_submitted" (underscore instead of
1293 hyphen). Since every message generated by Exim is necessarily in response
1294 to another message, thes have all been changed to:
1296 Auto-Submitted: auto-replied
1298 in accordance with these statements in the RFC:
1300 The auto-replied keyword:
1302 - SHOULD be used on messages sent in direct response to another
1303 message by an automatic process,
1305 - MUST NOT be used on manually-generated messages,
1307 - MAY be used on Delivery Status Notifications (DSNs) and Message
1308 Disposition Notifications (MDNs),
1310 - MUST NOT be used on messages generated by automatic or periodic
1311 processes, except for messages which are automatic responses to
1314 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1315 to the default Received: header definition.
1317 PH/12 Added log selector acl_warn_skipped (default on).
1319 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1320 variables because (a) they are in the wrong storage pool and (b) even if
1321 they were copied, it wouldn't work properly because of the caching.
1323 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1324 checking when decoding. Apparently there are clients that generate
1325 overlong encoded strings. Why am I not surprised?
1327 PH/15 If the first argument of "${if match_address" was not empty, but did not
1328 contain an "@" character, Exim crashed. Now it writes a panic log message
1329 and treats the condition as false.
1331 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1333 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1334 extension "envelope-auth". The code is finished and in agreement with
1335 other implementations, but there is no documentation so far and in fact,
1336 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1337 not changing the active code.
1339 Print executed "if" and "elsif" statements when debugging is used. This
1340 helps a great deal to understand what a filter does.
1342 Document more things not specified clearly in RFC3028. I had all this
1343 sorted out, when out of a sudden new issues came to my mind. Oops."
1345 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1348 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1349 canonical form (as documented). However, after a host name lookup from
1350 the IP address, check_host() was doing a simple string comparison with
1351 addresses acquired from the DNS when checking that the found name did
1352 have the original IP as one of its addresses. Since any found IPv6
1353 addresses are likely to be in abbreviated form, the comparison could
1354 fail. Luckily, there already exists a function for doing the comparison
1355 by converting both addresses to binary, so now that is used instead of
1356 the text comparison.
1358 PH/20 There was another similar case to PH/19, when a complete host name was
1359 given in a host list; looking up its IP address could give an abbreviated
1360 form, whereas the current host's name might or might not be abbreviated.
1361 The same fix has been applied.
1367 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1368 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1371 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1372 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1374 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1376 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1377 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1378 Also, for "Auto-Submitted", treat anything other than "no" as
1379 non-personal, in accordance with RFC 3834. (Previously it treated
1380 anything starting "auto-" as non-personal.)
1382 TF/01 The control=submission/name=... option had a problem with syntax
1383 errors if the name included a slash character. The /name= option
1384 now slurps the rest of the string, so it can include any characters
1385 but it must come last in the list of options (after /sender_retain
1388 PH/05 Some modifications to the interface to the fake nameserver for the new
1396 TK/01 Added the "success_on_redirect" address verification option. See
1397 NewStuff for rationale and an example.
1399 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1401 PH/02 Patch to exigrep to allow it to work on syslog lines.
1403 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1404 fread() to skip over the body file's header line, because in Cygwin the
1405 header line is locked and is inaccessible.
1407 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1408 co-exist for some time) to make it clear that it is the Exim ID that is
1409 referenced, not the Message-ID: header line.
1411 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1412 string_format() function, because snprintf() does not exist on all
1415 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1416 ${stat: expansion item.
1418 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1419 protocol synchronization error", to keep the pedants happy.
1421 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1422 well as for IRIX systems, when gcc is being used. See the host.c source
1425 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1427 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1430 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1431 transport and to the smtp transport in LMTP mode.
1433 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1435 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1436 run a filter in a subprocess. This could lead to confusion in subsequent
1437 lookups in the parent process. There should also be a search_tidyup() at
1438 the end of the subprocess.
1440 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1441 only if the host matched helo_try_verify_hosts, which caused the
1442 verification to occur when the EHLO/HELO command was issued. The ACL just
1443 tested the remembered result. Now, if a previous verification attempt has
1444 not happened, "verify = helo" does it there and then.
1446 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1448 TK/03 Fix log output including CR from clamd.
1450 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1451 exist provoked a memory error which could cause a segfault.
1453 PH/15 Installed PCRE 6.2
1455 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1457 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1458 of the problem. Specifically, suggested +O2 rather than +O1 for the
1461 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1463 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1464 record, the port was not correctly logged when the outgoing_port log
1465 selector was set (it logged the transort's default port).
1467 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1468 fallback_hosts, and "hosts" in the smtp transport.
1470 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1471 host errors such as "Connection refused".
1473 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1474 authentication with radiusclient 0.4.9:
1476 - Error returned from rc_read_config was caught wrongly
1477 - Username/password not passed on to radius server due to wrong length.
1479 The presumption is that some radiusclient API changes for 4.51/PH/17
1480 were not taken care of correctly. The code is still untested by me (my
1481 Linux distribution still has 0.3.2 of radiusclient), but it was
1482 contributed by a Radius user.
1484 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1485 expanding the "port" option of the smtp transport.
1487 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1488 while reading a MIME header. Thanks to Tom Hughes for a patch.
1490 PH/24 Include config.h inside local_scan.h so that configuration settings are
1493 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1494 that in an ACL for RCPT (for example), you can examine exactly what was
1497 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1498 commands, but it was not correctly comparing the address with the actual
1499 client host address. Thus, it would show the EHLO address in Received:
1500 header lines when this was not necessary.
1502 PH/27 Added the % operator to ${eval:}.
1504 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1505 it should be ignoring failures (because with -C, for example, it has lost
1506 privilege). It wasn't ignoring creation failures other than "already
1509 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1512 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1513 because an input line was too long, either on its own, or by virtue of
1514 too many continuations, the temporary file was not being removed, and the
1515 return code was incorrect.
1517 PH/31 Missing "BOOL" in function definition in filtertest.c.
1519 PH/32 Applied Sieve patches from the maintainer.
1521 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1523 PH/33 Added "verify = not_blind".
1525 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1526 Local/Makefile (with some defaults set). These are used in built scripts
1527 such as exicyclog, but they have never been used in the exim_install
1528 script (though there are many overriding facilities there). I have
1529 arranged that the exim_install script now takes note of these two
1532 PH/35 Installed configuration files for Dragonfly.
1534 PH/36 When a locally submitted message by a trusted user did not contain a
1535 From: header, and the sender address was obtained from -f or from an SMTP
1536 MAIL command, and the trusted user did not use -F to supply a sender
1537 name, $originator_name was incorrectly used when constructing a From:
1538 header. Furthermore, $originator_name was used for submission mode
1539 messages from external hosts without From: headers in a similar way,
1540 which is clearly wrong.
1542 PH/37 Added control=suppress_local_fixups.
1544 PH/38 When log_selector = +received_sender was set, and the addition of the
1545 sender made the log line's construction buffer exactly full, or one byte
1546 less than full, an overflow happened when the terminating "\n" was
1549 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1550 when the result of a list match is failure because a DNS lookup failed.
1552 PH/40 RM_COMMAND is now used in the building process.
1554 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1555 the "build-* directories that it finds.
1557 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1558 address in a domain literal was a prefix of an interface address.
1560 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1561 when verifying a sender address, unless rewrite_headers is false.
1563 PH/44 Wrote a long comment about why errors_to addresses are verified as
1564 recipients, not senders.
1566 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1567 the ratelimit ACL was added.
1569 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1571 PH/46 Added extra information about PostgreSQL errors to the error string.
1573 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1574 avoiding the need to install special zones in a real server. This is
1575 backwards compatible; if it can't find the fake resolver, it drops back.
1576 Thus, both old and new test suites can be run.
1578 TF/02 Added util/ratelimit.pl
1580 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1581 clock is set back in time.
1583 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1584 Candler <B.Candler@pobox.com>.
1586 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1587 OK for addresses that are the result of redirections.
1589 PH/48 A number of further additions for the benefit of the new test suite,
1590 including a fake gethostbyname() that interfaces to the fake DNS resolver
1593 TF/06 The fix for widen_domains has also been applied to qualify_single and
1594 search_parents which are the other dnslookup options that can cause
1597 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1600 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1602 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1603 service" if it encountered an MX record with an empty target hostname.
1604 The message is now "an MX or SRV record indicated no SMTP service".
1606 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1607 if the DNS of the sending site is misconfigured. This is quite a
1608 common situation. This change restores the behaviour of treating a
1609 helo verification defer as a failure.
1611 PH/51 If self=fail was set on a router, the bounce message did not include the
1612 actual error message.
1618 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1620 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1621 command itself ended in error, the underlying message about the transport
1622 filter timeout was being overwritten with the pipe command error. Now the
1623 underlying error message should be appended to the second error message.
1625 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1627 PH/02 Reduce the amount of output that "make" produces by default. Full output
1628 can still be requested.
1630 PH/03 The warning log line about a condition test deferring for a "warn" verb
1631 was being output only once per connection, rather than after each
1632 occurrence (because it was using the same function as for successful
1633 "warn" verbs). This seems wrong, so I have changed it.
1635 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1636 it should not have, which might have caused a crash in the right
1637 circumstances, but probably never did.
1639 PH/04 Installed a modified version of Tony Finch's patch to make submission
1640 mode fix the return path as well as the Sender: header line, and to
1641 add a /name= option so that you can make the user's friendly name appear
1644 TF/03 Added the control = fakedefer ACL modifier.
1646 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1647 Mark Lowes for thorough testing.
1649 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1651 TK/03 Merged latest SRS patch from Miles Wilton.
1653 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1654 with the definition in sysexits.h (which is #included earlier).
1655 Fortunately, Exim does not actually use EX_OK. The code used to try to
1656 preserve the sysexits.h value, by assumimg that macro definitions were
1657 scanned for macro replacements. I have been disabused of this notion,
1658 so now the code just undefines EX_OK before #including unistd.h.
1660 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1661 in the smtp transport. When a block could not be written in a single
1662 write() function, the timeout was being re-applied to each part-write.
1663 This seems wrong - if the receiver was accepting one byte at a time it
1664 would take for ever. The timeout is now adjusted when this happens. It
1665 doesn't have to be particularly precise.
1667 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1668 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1670 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1671 without a domain if the check to <postmaster@domain> fails.
1673 SC/01 Eximstats: added -xls and the ability to specify output files
1674 (patch written by Frank Heydlauf).
1676 SC/02 Eximstats: use FileHandles for outputing results.
1678 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1680 SC/04 Eximstats: fixed display of large numbers with -nvr option
1682 SC/05 Eximstats: fixed merging of reports with empty tables.
1684 SC/06 Eximstats: added the -include_original_destination flag
1686 SC/07 Eximstats: removed tabs and trailing whitespace.
1688 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1690 TK/06 MBOX spool code: Add real "From " MBOX separator line
1691 so the .eml file is really in mbox format (even though
1692 most programs do not really care). Patch from Alex Miller.
1694 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1695 The latter is generated from $received_to and is only set if the
1696 message has one envelope recipient. SA can use these headers,
1697 obviously out-of-the-box. Patch from Alex Miller.
1699 PH/08 The ${def test on a variable was returning false if the variable's
1700 value was "0", contrary to what the specification has always said!
1701 The result should be true unless the variable is empty.
1703 PH/09 The syntax error of a character other than { following "${if
1704 def:variable_name" (after optional whitespace) was not being diagnosed.
1705 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1706 accidental colon was present, for example, could give incorrect results.
1708 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1709 result is used (not including appendfile, where other changes are about
1712 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1713 This involved changing a lot of size variables from int to off_t. It
1714 should work with maildirs and everything.
1716 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1717 spamd dying while we are connected to it.
1719 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1722 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1723 function for BDB 4.3.
1725 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1727 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1728 the sources for PCRE have been re-arranged and more files are now
1731 PH/15 The code I had for printing potentially long long variables in PH/11
1732 above was not the best (it lost precision). The length of off_t variables
1733 is now inspected at build time, and an appropriate printing format (%ld
1734 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1735 to be "long long int" or "long int". This is needed for the internal
1736 formatting function string_vformat().
1738 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1739 the configuration file to be ":syslog", then the script "guesses" where
1740 the logs files are, rather than using the compiled in default. In our
1741 case the guess is not the same as the compiled default, so the script
1742 suddenly stopped working when I started to use syslog. The patch checks
1743 to see if log_file_path is "". If so, it attempts to read it from exim
1744 with no configuration file to get the compiled in version, before it
1745 falls back to the previous guessing code."
1747 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1748 implementing BATV in an Exim configuration. See NewStuff for the gory
1751 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1752 Makefile that are specific to HP-UX.
1754 PH/18 If the "use_postmaster" option was set for a recipient callout together
1755 with the "random" option, the postmaster address was used as the MAIL
1756 FROM address for the random test, but not for the subsequent recipient
1757 test. It is now used for both.
1759 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1760 patch removes a few documentation additions to RFC 3028, because the
1761 latest draft now contains them. It adds the new en;ascii-case comparator
1762 and a new error check for 8bit text in MIME parts. Comparator and
1763 require names are now matched exactly. I enabled the subaddress
1764 extension, but it is not well tested yet (read: it works for me)."
1766 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1767 rework some of the code of TK/09 above to avoid the hardwired use of
1768 "%lld" and "long long". Replaced the call to snprintf() with a call to
1771 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1772 records point to non-existent hosts", "retry timeout exceeded", and
1773 "retry time not reached for any host after a long failure period".
1775 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1776 experimental DomainKeys support:
1778 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1779 (2) On an error such as an illegally used "control", the wrong name for
1780 the control was given.
1782 These problems did NOT occur unless DomainKeys support was compiled.
1784 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1786 PH/24 Added ${if match_ip condition.
1788 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1789 databases so that it will be absolutely obvious if a crash occurs in the
1790 DB library. This is a regular occurrence (often caused by mis-matched
1793 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1794 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1795 picked up on a user's system that detects such things. There doesn't seem
1796 to be a gcc warning option for this - only an attribute that has to be
1797 put on the function's prototype. It seems that in Fedora Core 4 they have
1798 set this on a number of new functions. No doubt there will be more in due
1801 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1802 specify a transport. However, if an address that was verified by such a
1803 router was the subject of a callout, Exim crashed because it tried to
1804 read the rcpt_include_affixes from the non-existent transport. Now it
1805 just assumes that the setting of that option is false. This bug was
1806 introduced by 4.51/PH/31.
1808 PH/28 Changed -d+all to exclude +memory, because that information is very
1809 rarely of interest, but it makes the output a lot bigger. People tend to
1810 do -d+all out of habit.
1812 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1813 code in os-type was giving problems when libc.so lives in lib64, like on
1816 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1817 aren't the modern standard, and it seems that some systems' include files
1818 don't always have them. Exim was already checking for some of the newer
1819 ones like T_AAAA, and defining it itself. I've added checks for all the
1820 record types that Exim uses.
1822 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1823 not automatically generating a new one, as it is supposed to. This
1824 prevented TLS from working. If the file did exist, but contained invalid
1825 data, a new version was generated, as expected. It was only the case of a
1826 non-existent file that was broken.
1828 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1829 with a change in libdomainkeys > 0.64.
1831 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1832 from DNS. If the selector record carries the flag, it now has
1833 precedence over the domain-wide flag.
1835 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1837 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1838 the use of a port name that isn't defined in /etc/services) occurred, the
1839 message was deferred as in a normal delivery, and thus remained on the
1840 spool, instead of being failed because of the mua_wrapper setting. This
1841 is now fixed, and I tidied up some of the mua_wrapper messages at the
1844 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1845 the messages in a hash of arrays rather than using individual hashes.
1846 This is a bit cleaner and results in dramatic memory savings, albeit
1847 at a slight CPU cost.
1849 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1850 as requested by Marc Sherman.
1852 SC/10 Eximstats: added histograms for user specified patterns as requested
1855 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1857 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1858 fopen() in the content-scanning modules that did not already have it.
1864 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1865 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1867 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1869 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1871 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1872 to test Sieve filters that use "vacation".
1874 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1875 that changes the way the GnuTLS parameters are stored in the cache file.
1876 The new format can be generated externally. For backward compatibility,
1877 if the data in the cache doesn't make sense, Exim assumes it has read an
1878 old-format file, and it generates new data and writes a new file. This
1879 means that you can't go back to an older release without removing the
1882 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1883 work if there are any delivery delays because "one_time" forces the
1884 parent to be marked "delivered", so its unseen clone is never tried
1885 again. For this reason, Exim now forbids the simultaneous setting of
1888 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1889 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1890 entry if you want to know the details. The fix, however, neglected to
1891 consider the case when local delivery batching is involved. The test for
1892 "previously delivered" was not happening when checking to see if an
1893 address could be batched with a previous (undelivered) one; under
1894 certain circumstances this could lead to multiple deliveries to the same
1897 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1898 in its include files, and this causes problems building Exim.
1900 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1901 header_syntax) but Exim was just ignoring anything given after a slash.
1902 In particular, this caused confusion with an attempt to use "verify =
1903 reverse_host_lookup/defer_ok". An error is now given when options are
1904 supplied for verify items that do not have them. (Maybe reverse_host_
1905 lookup should have a defer_ok option, but that's a different point.)
1907 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1908 defined by RFC 821) to 2048, because there were problems with some AUTH
1909 commands, and RFC 1869 says the size should be increased for extended
1910 SMTP commands that take arguments.
1912 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1915 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1916 "unknown" error; now it says that the functionality isn't in the binary.
1918 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1919 an address' error message when a string expansion fails (syntax or
1920 whatever). Otherwise the password may appear in the log. Following change
1921 PH/42 below, there is no longer a chance of it appearing in a bounce
1924 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1926 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1927 of its addresses was ever tried. (Bugzilla bug #2.)
1929 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1930 the result incorrectly in the debug output. (It correctly added a newline
1931 to what was transported.)
1933 TF/01 Added $received_time.
1935 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1936 commented out examples of how to interface to a virus scanner and to
1937 SpamAssassin. Also added commented examples of av_scanner and
1938 spamd_address settings.
1940 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1941 and controls are allowed in which ACLs. There were a couple of minor
1942 errors. Some of the entries in the conditions table (which is a table of
1943 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1944 negation of where the condition IS allowed.
1946 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1948 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1949 header file does not have a version number, so I've had to invent a new
1950 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1951 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1952 radiusclient), but it was contributed by a Radius user.
1954 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1955 files or directories, for OpenSSL.
1957 PH/19 When an Exim process that is running as root has to create an Exim log
1958 file, it does so in a subprocess that runs as exim:exim so as to get the
1959 ownership right at creation (otherwise, other Exim processes might see
1960 the file with the wrong ownership). There was no test for failure of this
1961 fork() call, which would lead to the process getting stuck as it waited
1962 for a non-existent subprocess. Forks do occasionally fail when resources
1963 run out. I reviewed all the other calls to fork(); they all seem to check
1966 PH/20 When checking for unexpected SMTP input at connect time (before writing
1967 the banner), Exim was not dealing correctly with a non-positive return
1968 from the read() function. If the client had disconnected by this time,
1969 the result was a log entry for a synchronization error with an empty
1970 string after "input=" when read() returned zero. If read() returned -1
1971 (an event I could not check), uninitialized data bytes were printed.
1972 There were reports of junk text (parts of files, etc) appearing after
1975 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1977 PH/22 Added support for macro redefinition, and (re)definition in between
1978 driver and ACL definitions.
1980 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1981 forgetting to use the resulting value; it was using the unexpanded value.
1983 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1984 hadn't been configured. The fix is from Juergen Kreileder, who
1985 understands it better than I do:
1987 "Here's what I see happening with three configured cyrus_sasl
1988 authenticators configured (plain, login, cram-md5):
1990 On startup auth_cyrus_sasl_init() gets called for each of these.
1991 This means three calls to sasl_listmech() without a specified mech_list.
1992 => SASL tests which mechs of all available mechs actually work
1993 => three warnings about OTP not working
1994 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1996 With the patch, sasl_listmech() also gets called three times. But now
1997 SASL's mech_list option is set to the server_mech specified in the the
1998 authenticator. Or in other words, the answer from sasl_listmech()
1999 gets limited to just the mech you're testing for (which is different
2001 => the return list contains just 'plain' or 'login', 'cram-md5' or
2002 nothing depending on the value of ob->server_mech.
2004 I've just tested the patch: Authentication still works fine,
2005 unavailable mechs specified in the exim configuration are still
2006 caught, and the auth.log warnings about OTP are gone."
2008 PH/25 When debugging is enabled, the contents of the command line are added
2009 to the debugging output, even when log_selector=+arguments is not
2012 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2013 answer is "GNU", and only if the return is "GNU/something" is the answer
2016 PH/27 $acl_verify_message is now set immediately after the failure of a
2017 verification in an ACL, and so is available in subsequent modifiers. In
2018 particular, the message can be preserved by coding like this:
2020 warn !verify = sender
2021 set acl_m0 = $acl_verify_message
2023 Previously, $acl_verify_message was set only while expanding "message"
2024 and "log_message" when a very denied access.
2026 PH/28 Modified OS/os.c-Linux with
2028 -#ifndef OS_LOAD_AVERAGE
2029 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2031 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2032 nomenclature these days.)
2034 PH/29 Installed patch from the Sieve maintainer that adds the options
2035 sieve_useraddress and sieve_subaddress to the redirect router.
2037 PH/30 In these circumstances:
2038 . Two addresses routed to the same list of hosts;
2039 . First host does not offer TLS;
2040 . First host accepts first address;
2041 . First host gives temporary error to second address;
2042 . Second host offers TLS and a TLS session is established;
2043 . Second host accepts second address.
2044 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2045 and peerdn, if requested) that were in fact used only for the second
2048 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2049 attention to any local part prefix or suffix that was matched by the
2050 router that accepted the address. It now behaves in the same way as it
2051 does for delivery: the affixes are removed from the local part unless
2052 rcpt_include_affixes is set on the transport.
2054 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2055 timeout during the DATA phase of an incoming message.
2057 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2058 applied a patch sanctioned by the Sieve maintainer.
2060 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2061 the uid or gid is negative. A case of a negative gid caused this to be
2062 noticed. The fix allows for either to be negative.
2064 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2065 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2068 PH/36 The change PH/12 above was broken. Fixed it.
2070 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2071 the grounds that routing the same address twice would always produce the
2072 same answer. This might have been true once, but it is certainly no
2073 longer true now. Routing a child address may depend on the previous
2074 routing that produced that child. Some complicated redirection strategies
2075 went wrong when messages had multiple recipients, and made Exim's
2076 behaviour dependent on the order in which the addresses were given.
2078 I have moved the duplicate checking until after the routing is complete.
2079 Exim scans the addresses that are assigned to local and remote
2080 transports, and removes any duplicates. This means that more work will be
2081 done, as duplicates will always all be routed, but duplicates are
2082 presumably rare, so I don't expect this is of any significance.
2084 For deliveries to pipes, files, and autoreplies, the duplicate checking
2085 still happens during the routing process, since they are not going to be
2088 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2089 It corrects a timeout issue with spamd. This is Ian's comment: "The
2090 background is that sometimes spamd either never reads data from a
2091 connection it has accepted, or it never writes response data. The exiscan
2092 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2093 blindly assumes that writes won't block so it may never time out."
2095 PH/39 Allow G after quota size as well as K and M.
2097 PH/40 The value set for $authenticated_id in an authenticator may not contain
2098 binary zeroes or newlines because the value is written to log lines and
2099 to spool files. There was no check on this. Now the value is run through
2100 the string_printing() function so that such characters are converted to
2101 printable escape sequences.
2103 PH/41 $message_linecount is a new variable that contains the total number of
2104 lines in the message. Compare $body_linecount, which is the count for the
2107 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2108 bounce and delay warning messages, except in certain special cases, which
2111 (a) An SMTP error message from a remote host;
2112 (b) A message specified in a :fail: redirection;
2113 (c) A message specified in a "fail" command in a system filter;
2114 (d) A message specified in a FAIL return from the queryprogram router;
2115 (e) A message specified by the cannot_route_message router option.
2117 In these cases only, Exim does include the error details in bounce and
2118 warning messages. There are also a few cases where bland messages such
2119 as "unrouteable address" or "local delivery error" are given.
2121 PH/43 $value is now also set for the "else" part of a ${run expansion.
2123 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2124 being worked on, but at least Exim now implements the latest version to
2127 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2128 process to complete was treated as a delivery failure, a timeout while
2129 writing the message to the pipe was logged, but erroneously treated as a
2130 successful delivery. Such timeouts include transport filter timeouts. For
2131 consistency with the overall process timeout, these timeouts are now
2132 treated as errors, giving rise to delivery failures by default. However,
2133 there is now a new Boolean option for the pipe transport called
2134 timeout_defer, which, if set TRUE, converts the failures into defers for
2135 both kinds of timeout. A transport filter timeout is now identified in
2138 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2139 systems where "make" and "gmake" are different, calling "gmake" at top
2140 level broke things. I've arranged for the value of $(MAKE) to be passed
2141 from the Makefile to this script so that it can call the same version of
2145 A note about Exim versions 4.44 and 4.50
2146 ----------------------------------------
2148 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2149 changes of various kinds. As a consequence, a big documentation update was
2150 needed. This delayed the release for rather longer than seemed good, especially
2151 in the light of a couple of (minor) security issues. Therefore, the changes
2152 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2153 release. So 4.44 and 4.50 are in effect two different branches that both start
2156 I have left the 4.50 change log unchanged; it contains all the changes since
2157 4.43. The change log for 4.44 is below; many of its items are identical to
2158 those for 4.50. This seems to be the most sensible way to preserve the
2159 historical information.
2165 1. Minor wording change to the doc/README.SIEVE file.
2167 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2168 computation of the current number of files was incorrect.
2170 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2171 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2174 4. Give more explanation in the error message when the command for a transport
2175 filter fails to execute.
2177 5. There are several places where Exim runs a non-Exim command in a
2178 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2179 was being done only for the command run by the queryprogram router. It is
2180 now done for all such subprocesses. The other cases are: ${run, transport
2181 filters, and the commands run by the lmtp and pipe transports.
2183 6. Added CONFIGURE_GROUP build-time option.
2185 7. Some older OS have a limit of 256 on the maximum number of file
2186 descriptors. Exim was using setrlimit() to set 1000 as a large value
2187 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2188 systems. I've change it so that if it can't get 1000, it tries for 256.
2190 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2191 was an oversight, and furthermore, ever since the addition of extra
2192 controls (e.g. 4.43/32), the checks on when to allow different forms of
2193 "control" were broken. There should now be diagnostics for all cases when a
2194 control that does not make sense is encountered.
2196 9. Added the /retain_sender option to "control=submission".
2198 10. $recipients is now available in the predata ACL (oversight).
2200 11. Tidy the search cache before the fork to do a delivery from a message
2201 received from the command line. Otherwise the child will trigger a lookup
2202 failure and thereby defer the delivery if it tries to use (for example) a
2203 cached ldap connection that the parent has called unbind on.
2205 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2206 of $address_data from the recipient verification was clobbered by the
2207 sender verification.
2209 13. The value of address_data from a sender verification is now available in
2210 $sender_address_data in subsequent conditions in the ACL statement.
2212 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2214 15. Added a new option "connect=<time>" to callout options, to set a different
2217 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2218 was its contents. (It was OK if the option was not defined at all.)
2220 17. A "Completed" log line is now written for messages that are removed from
2221 the spool by the -Mrm option.
2223 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2224 information about exactly what failed.
2226 19. Added -dd to debug only the daemon process.
2228 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2229 handles timeouts, both on the server side and network timeouts. Renamed the
2230 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2232 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2233 It is now set to "smtps".
2235 22. $host_address is now set to the target address during the checking of
2236 ignore_target_hosts.
2238 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2239 being passed; this would have caused $sender_host_name to have been used if
2240 matching the list had actually called for a host name (not very likely,
2241 since this list is usually IP addresses). A host name is now passed as
2244 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2245 code that specifies a non-restarting handler (typically sigaction() in
2246 modern systems) in an attempt to fix a rare and obscure crash bug.
2248 25. Narrowed the window for a race in the daemon that could cause it to ignore
2249 SIGCHLD signals. This is not a major problem, because they are used only to
2250 wake it up if nothing else does.
2252 26. A malformed maildirsize file could cause Exim to calculate negative values
2253 for the mailbox size or file count. Odd effects could occur as a result.
2254 The maildirsize information is now recalculated if the size or filecount
2257 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2258 support for a long time. Removed HAVE_SYS_VFS_H.
2260 28. Installed the latest version of exipick from John Jetmore.
2262 29. In an address list, if the pattern was not a regular expression, an empty
2263 subject address (from a bounce message) matched only if the pattern was an
2264 empty string. Non-empty patterns were not even tested. This was the wrong
2265 because it is perfectly reasonable to use an empty address as part of a
2266 database query. An empty address is now tested by patterns that are
2267 lookups. However, all the other forms of pattern expect the subject to
2268 contain a local part and a domain, and therefore, for them, an empty
2269 address still always fails if the pattern is not itself empty.
2271 30. Exim went into a mad DNS loop when attempting to do a callout where the
2272 host was specified on an smtp transport, and looking it up yielded more
2273 than one IP address.
2275 31. Re-factored the code for checking spool and log partition space into a
2276 function that finds that data and another that does the check. The former
2277 is then used to implement four new variables: $spool_space, $log_space,
2278 $spool_inodes, and $log_inodes.
2280 32. The RFC2047 encoding function was originally intended for short strings
2281 such as real names; it was not keeping to the 75-character limit for
2282 encoded words that the RFC imposes. It now respects the limit, and
2283 generates multiple encoded words if necessary. To be on the safe side, I
2284 have increased the buffer size for the ${rfc2047: expansion operator from
2287 33. It is now permitted to omit both strings after an "if" condition; if the
2288 condition is true, the result is "true". As before, when the second string
2289 is omitted, a false condition yields an empty string. This makes it less
2290 cumbersome to write custom ACL and router conditions.
2292 34. Failure to deliver a bounce message always caused it to be frozen, even if
2293 there was an errors_to setting on the router. The errors_to setting is now
2296 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2297 canonical form (fully expanded) before being placed in
2298 $sender_host_address.
2300 36. The table in the code that translates DNS record types into text (T_A to
2301 "A" for instance) was missing entries for NS and CNAME. It is just possible
2302 that this could have caused confusion if both these types were looked up
2303 for the same domain, because the text type is used as part of Exim's
2304 per-process caching. But the chance of anyone hitting this buglet seems
2307 37. The dnsdb lookup has been extended in a number of ways.
2309 (1) There is a new type, "zns", which walks up the domain tree until it
2310 finds some nameserver records. It should be used with care.
2312 (2) There is a new type, "mxh", which is like "mx" except that it returns
2313 just the host names, not the priorities.
2315 (3) It is now possible to give a list of domains (or IP addresses) to be
2316 looked up. The behaviour when one of the lookups defers can be
2317 controlled by a keyword.
2319 (4) It is now possible to specify the separator character for use when
2320 multiple records are returned.
2322 38. The dnslists ACL condition has been extended: it is now possible to supply
2323 a list of IP addresses and/or domains to be looked up in a particular DNS
2326 39. Added log_selector=+queue_time_overall.
2328 40. When running the queue in the test harness, wait just a tad after forking a
2329 delivery process, to get repeatability of debugging output.
2331 41. Include certificate and key file names in error message when GnuTLS fails
2332 to set them up, because the GnuTLS error message doesn't include the name
2333 of the failing file when there is a problem reading it.
2335 42. Allow both -bf and -bF in the same test run.
2337 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2339 44. The "Exiscan patch" is now merged into the mainline Exim source.
2341 45. Sometimes the final signoff response after QUIT could fail to get
2342 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2343 < 0 before doing a fflush(). This bug looks as though it goes back to the
2344 introduction of TLS in release 3.20, but "sometimes" must have been rare
2345 because the tests only now provoked it.
2347 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2348 (this can affect the format of dates).
2350 47. exim_tidydb, when checking for the continued existence of a message for
2351 which it has found a message-specific retry record, was not finding
2352 messages that were in split spool directories. Consequently, it was
2353 deleting retry records that should have stayed in existence.
2355 48. Steve fixed some bugs in eximstats.
2357 49. The SPA authentication driver was not abandoning authentication and moving
2358 on to the next authenticator when an expansion was forced to fail,
2359 contradicting the general specification for all authenticators. Instead it
2360 was generating a temporary error. It now behaves as specified.
2362 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2363 (the order specifies the preference for clients). The order is now AES256,
2364 AES128, 3DES, ARCFOUR128.
2366 51. Small patch to Sieve code - explicitly set From: when generating an
2369 52. Exim crashed if a remote delivery caused a very long error message to be
2370 recorded - for instance if somebody sent an entire SpamAssassin report back
2371 as a large number of 550 error lines. This bug was coincidentally fixed by
2372 increasing the size of one of Exim's internal buffers (big_buffer) that
2373 happened as part of the Exiscan merge. However, to be on the safe side, I
2374 have made the code more robust (and fixed the comments that describe what
2377 53. Now that there can be additional text after "Completed" in log lines (if
2378 the queue_time_overall log selector is set), a one-byte patch to exigrep
2379 was needed to allow it to recognize "Completed" as not the last thing in
2382 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2383 patch that reportedly fixes this has been added. I am not expert enough to
2384 create a test for it. This is what the patch creator wrote:
2386 "I found a little strange behaviour of ldap code when working with
2387 Windows 2003 AD Domain, where users was placed in more than one
2388 Organization Units. When I tried to give exim partial DN, the exit code
2389 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2390 But simultaneously result of request was absolutely normal ldap result,
2391 so I produce this patch..."
2393 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2394 REFERENCE, so I have modified the code to exclude the patch when that macro
2397 55. Some experimental protocols are using DNS PTR records for new purposes. The
2398 keys for these records are domain names, not reversed IP addresses. The
2399 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2400 leaves it alone. Component reversal etc. now happens only for IP addresses.
2403 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2405 57. Double the size of the debug message buffer (to 2048) so that more of very
2406 long debug lines gets shown.
2408 58. The exicyclog utility now does better if the number of log files to keep
2409 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2411 59. Two changes related to the smtp_active_hostname option:
2413 (1) $smtp_active_hostname is now available as a variable.
2414 (2) The default for smtp_banner uses $smtp_active_hostname instead
2415 of $primary_hostname.
2417 60. The host_aton() function is supposed to be passed a string that is known
2418 to be a valid IP address. However, in the case of IPv6 addresses, it was
2419 not checking this. This is a hostage to fortune. Exim now panics and dies
2420 if the condition is not met. A case was found where this could be provoked
2421 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2422 components; fortuitously, this particular loophole had already been fixed
2423 by change 4.50/55 above.
2425 If there are any other similar loopholes, the new check in host_aton()
2426 itself should stop them being exploited. The report I received stated that
2427 data on the command line could provoke the exploit when Exim was running as
2428 exim, but did not say which command line option was involved. All I could
2429 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2430 running as the user.
2433 61. There was a buffer overflow vulnerability in the SPA authentication code
2434 (which came originally from the Samba project). I have added a test to the
2435 spa_base64_to_bits() function which I hope fixes it.
2438 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2439 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2441 63. The daemon start-up calls getloadavg() while still root for those OS that
2442 need the first call to be done as root, but it missed one case: when
2443 deliver_queue_load_max is set with deliver_drop_privilege. This is
2444 necessary for the benefit of the queue runner, because there is no re-exec
2445 when deliver_drop_privilege is set.
2447 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2448 This has been fixed.
2450 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2451 was in use, was not putting the data itself into the right store pool;
2452 consequently, it could be overwritten for a subsequent message in the same
2453 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2456 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2458 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2459 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2460 the function were treating the return as a boolean value, which happened to
2461 work because 0=false and not-0=true, but is not correct code.
2463 68. The host_aton() function was not handling scoped IPv6 addresses (those
2464 with, for example, "%eth0" on the end) correctly.
2466 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2467 negated items (that is, ~something) in unsigned ints. Some compilers
2468 apparently mutter when there is no cast.
2470 70. If an address verification called from an ACL failed, and did not produce a
2471 user-specific message (i.e. there was only a "system" message), nothing was
2472 put in $acl_verify_message. In this situation, it now puts the system
2475 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2476 session; change 4.31/43 added the unwanted input to the log line - except
2477 that it did not do this in the start of session case. It now does.
2479 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2480 This is wrong and can cause the other end to generate a synchronization
2481 error if it is another Exim or anything else that does the synchronization
2482 check. A QUIT command is no longer sent after a timeout.
2484 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2485 during host lookups.
2487 74. The defer_ok option of callout verification was not working if it was used
2488 when verifying addresses in header lines, that is, for this case:
2490 verify = header_sender/callout=defer_ok
2492 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2493 those file descriptors could be used for SMTP connections. If anything
2494 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2495 could be sent to the SMTP client, causing chaos. The daemon now opens
2496 stdin, stdout, and stderr to /dev/null when it puts itself into the
2499 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2500 log by default. The user can override this with suitable Perl magic.
2502 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2503 the log message when discard triggers, was not working for the DATA ACL or
2504 for the non-SMTP ACL.
2506 78. Error message wording change in sieve.c.
2508 79. If smtp_accept_max_per_host was set, the number of connections could be
2509 restricted to fewer than expected, because the daemon was trying to set up
2510 a new connection before checking whether the processes handling previous
2511 connections had finished. The check for completed processes is now done
2512 earlier. On busy systems, this bug wouldn't be noticed because something
2513 else would have woken the daemon, and it would have reaped the completed
2516 80. If a message was submitted locally by a user whose login name contained one
2517 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2518 It caused a spool format error. I have fixed the spool reading code. A
2519 related problem was that the "from" clause in the Received: line became
2520 illegal because of the space(s). It is now covered by ${quote_local_part.
2522 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2525 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2527 83. Patch from Sieve maintainer to fix unterminated string problem in
2528 "vacation" handling.
2530 84. Some minor changes to the Linux configuration files to help with other
2531 OS variants using glibc.
2533 85. One more patch for Sieve to update vacation handling to latest spec.
2536 ----------------------------------------------------
2537 See the note above about the 4.44 and 4.50 releases.
2538 ----------------------------------------------------
2544 1. Change 4.43/35 introduced a bug that caused file counts to be
2545 incorrectly computed when quota_filecount was set in an appendfile
2548 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2549 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2552 3. Give more explanation in the error message when the command for a transport
2553 filter fails to execute.
2555 4. There are several places where Exim runs a non-Exim command in a
2556 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2557 was being done only for the command run by the queryprogram router. It is
2558 now done for all such subprocesses. The other cases are: ${run, transport
2559 filters, and the commands run by the lmtp and pipe transports.
2561 5. Some older OS have a limit of 256 on the maximum number of file
2562 descriptors. Exim was using setrlimit() to set 1000 as a large value
2563 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2564 systems. I've change it so that if it can't get 1000, it tries for 256.
2566 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2567 was an oversight, and furthermore, ever since the addition of extra
2568 controls (e.g. 4.43/32), the checks on when to allow different forms of
2569 "control" were broken. There should now be diagnostics for all cases when a
2570 control that does not make sense is encountered.
2572 7. $recipients is now available in the predata ACL (oversight).
2574 8. Tidy the search cache before the fork to do a delivery from a message
2575 received from the command line. Otherwise the child will trigger a lookup
2576 failure and thereby defer the delivery if it tries to use (for example) a
2577 cached ldap connection that the parent has called unbind on.
2579 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2580 of $address_data from the recipient verification was clobbered by the
2581 sender verification.
2583 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2584 was its contents. (It was OK if the option was not defined at all.)
2586 11. A "Completed" log line is now written for messages that are removed from
2587 the spool by the -Mrm option.
2589 12. $host_address is now set to the target address during the checking of
2590 ignore_target_hosts.
2592 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2593 being passed; this would have caused $sender_host_name to have been used if
2594 matching the list had actually called for a host name (not very likely,
2595 since this list is usually IP addresses). A host name is now passed as
2598 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2599 code that specifies a non-restarting handler (typically sigaction() in
2600 modern systems) in an attempt to fix a rare and obscure crash bug.
2602 15. Narrowed the window for a race in the daemon that could cause it to ignore
2603 SIGCHLD signals. This is not a major problem, because they are used only to
2604 wake it up if nothing else does.
2606 16. A malformed maildirsize file could cause Exim to calculate negative values
2607 for the mailbox size or file count. Odd effects could occur as a result.
2608 The maildirsize information is now recalculated if the size or filecount
2611 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2612 support for a long time. Removed HAVE_SYS_VFS_H.
2614 18. Updated exipick to current release from John Jetmore.
2616 19. Allow an empty sender to be matched against a lookup in an address list.
2617 Previously the only cases considered were a regular expression, or an
2620 20. Exim went into a mad DNS lookup loop when doing a callout where the
2621 host was specified on the transport, if the DNS lookup yielded more than
2624 21. The RFC2047 encoding function was originally intended for short strings
2625 such as real names; it was not keeping to the 75-character limit for
2626 encoded words that the RFC imposes. It now respects the limit, and
2627 generates multiple encoded words if necessary. To be on the safe side, I
2628 have increased the buffer size for the ${rfc2047: expansion operator from
2631 22. Failure to deliver a bounce message always caused it to be frozen, even if
2632 there was an errors_to setting on the router. The errors_to setting is now
2635 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2636 canonical form (fully expanded) before being placed in
2637 $sender_host_address.
2639 24. Updated eximstats to version 1.33
2641 25. Include certificate and key file names in error message when GnuTLS fails
2642 to set them up, because the GnuTLS error message doesn't include the name
2643 of the failing file when there is a problem reading it.
2645 26. Expand error message when OpenSSL has problems setting up cert/key files.
2648 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2649 (this can affect the format of dates).
2651 28. exim_tidydb, when checking for the continued existence of a message for
2652 which it has found a message-specific retry record, was not finding
2653 messages that were in split spool directories. Consequently, it was
2654 deleting retry records that should have stayed in existence.
2656 29. eximstats updated to version 1.35
2657 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2658 1.35 - bugfix such that pie charts by volume are generated correctly
2660 30. The SPA authentication driver was not abandoning authentication and moving
2661 on to the next authenticator when an expansion was forced to fail,
2662 contradicting the general specification for all authenticators. Instead it
2663 was generating a temporary error. It now behaves as specified.
2665 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2666 (the order specifies the preference for clients). The order is now AES256,
2667 AES128, 3DES, ARCFOUR128.
2669 31. Small patch to Sieve code - explicitly set From: when generating an
2672 32. Exim crashed if a remote delivery caused a very long error message to be
2673 recorded - for instance if somebody sent an entire SpamAssassin report back
2674 as a large number of 550 error lines. This bug was coincidentally fixed by
2675 increasing the size of one of Exim's internal buffers (big_buffer) that
2676 happened as part of the Exiscan merge. However, to be on the safe side, I
2677 have made the code more robust (and fixed the comments that describe what
2680 33. Some experimental protocols are using DNS PTR records for new purposes. The
2681 keys for these records are domain names, not reversed IP addresses. The
2682 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2683 leaves it alone. Component reversal etc. now happens only for IP addresses.
2686 34. The host_aton() function is supposed to be passed a string that is known
2687 to be a valid IP address. However, in the case of IPv6 addresses, it was
2688 not checking this. This is a hostage to fortune. Exim now panics and dies
2689 if the condition is not met. A case was found where this could be provoked
2690 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2691 components; fortuitously, this particular loophole had already been fixed
2692 by change 4.50/55 or 4.44/33 above.
2694 If there are any other similar loopholes, the new check in host_aton()
2695 itself should stop them being exploited. The report I received stated that
2696 data on the command line could provoke the exploit when Exim was running as
2697 exim, but did not say which command line option was involved. All I could
2698 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2699 running as the user.
2702 35. There was a buffer overflow vulnerability in the SPA authentication code
2703 (which came originally from the Samba project). I have added a test to the
2704 spa_base64_to_bits() function which I hope fixes it.
2707 36. The daemon start-up calls getloadavg() while still root for those OS that
2708 need the first call to be done as root, but it missed one case: when
2709 deliver_queue_load_max is set with deliver_drop_privilege. This is
2710 necessary for the benefit of the queue runner, because there is no re-exec
2711 when deliver_drop_privilege is set.
2713 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2714 was in use, was not putting the data itself into the right store pool;
2715 consequently, it could be overwritten for a subsequent message in the same
2716 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2719 38. Sometimes the final signoff response after QUIT could fail to get
2720 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2721 < 0 before doing a fflush(). This bug looks as though it goes back to the
2722 introduction of TLS in release 3.20, but "sometimes" must have been rare
2723 because the tests only now provoked it.
2729 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2730 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2731 Now it returns an integer. A number of calls were still expecting a T/F
2732 return. Fortuitously, in all cases, the tests worked in OK situations,
2733 which is the norm. However, things would have gone wrong on any write
2734 failures on the smtp file descriptor. This function is used when sending
2735 messages over SMTP and also when doing verify callouts.
2737 2. When Exim is called to do synchronous delivery of a locally submitted
2738 message (the -odf or -odi options), it no longer closes stderr before doing
2741 3. Implemented the mua_wrapper option.
2743 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2745 5. Implemented the functions header_remove(), header_testname(),
2746 header_add_at_position(), and receive_remove_recipient(), and exported them
2749 6. If an ACL "warn" statement specified the addition of headers, Exim already
2750 inserted X-ACL-Warn: at the start if there was no header name. However, it
2751 was not making this test for the second and subsequent header lines if
2752 there were newlines in the string. This meant that an invalid header could
2753 be inserted if Exim was badly configured.
2755 7. Allow an ACL "warn" statement to add header lines at the start or after all
2756 the Received: headers, as well as at the end.
2758 8. Added the rcpt_4xx retry error code.
2760 9. Added postmaster_mailfrom=xxx to callout verification option.
2762 10. Added mailfrom=xxxx to the callout verification option, for verify=
2765 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2766 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2768 12. Inserted some casts to stop certain compilers warning when using pointer
2769 differences as field lengths or precisions in printf-type calls (mostly
2770 affecting debugging statements).
2772 13. Added optional readline() support for -be (dynamically loaded).
2774 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2775 same clock tick as a message's arrival, so that its received time was the
2776 same as the "first fail" time on the retry record, and that message
2777 remained on the queue past the ultimate address timeout, every queue runner
2778 would try a delivery (because it was past the ultimate address timeout) but
2779 after another failure, the ultimate address timeout, which should have then
2780 bounced the address, did not kick in. This was a "< instead of <=" error;
2781 in most cases the first failure would have been in the next clock tick
2782 after the received time, and all would be well.
2784 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2785 being recognized when the domain list was tested by the match_domain
2786 condition in an expansion string.
2788 16. Added the ${str2b64: operator.
2790 17. Exim was always calling setrlimit() to set a large limit for the number of
2791 processes, without checking whether the existing limit was already
2792 adequate. (It did check for the limit on file descriptors.) Furthermore,
2793 errors from getrlimit() and setrlimit() were being ignored. Now they are
2794 logged to the main and panic logs, but Exim does carry on, to try to do its
2795 job under whatever limits there are.
2797 18. Imported PCRE 5.0.
2799 19. Trivial typo in log message " temporarily refused connection" (the leading
2802 20. If the log selector return_path_on_delivery was set and an address was
2803 redirected to /dev/null, the delivery process crashed because it assumed
2804 that a return path would always be set for a "successful" delivery. In this
2805 case, the whole delivery is bypassed as an optimization, and therefore no
2808 21. Internal re-arrangement: the function for sending a challenge and reading
2809 a response while authentication was assuming a zero-terminated challenge
2810 string. It's now changed to take a pointer and a length, to allow for
2811 binary data in such strings.
2813 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2815 23. Exim was not respecting finduser_retries when seeking the login of the
2816 uid under which it was called; it was always trying 10 times. (The default
2817 setting of finduser_retries is zero.) Also, it was sleeping after the final
2818 failure, which is pointless.
2820 24. Implemented tls_on_connect_ports.
2822 25. Implemented acl_smtp_predata.
2824 26. If the domain in control=submission is set empty, Exim assumes that the
2825 authenticated id is a complete email address when it generates From: or
2826 Sender: header lines.
2828 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2829 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2830 chown and chgrp in /bin and hostname in /usr/bin.
2832 28. Exim was keeping the "process log" file open after each use, just as it
2833 does for the main log. This opens the possibility of it remaining open for
2834 long periods when the USR1 signal hits a daemon. Occasional processlog
2835 errors were reported, that could have been caused by this. Anyway, it seems
2836 much more sensible not to leave this file open at all, so that is what now
2839 29. The long-running daemon process does not normally write to the log once it
2840 has entered its main loop, and it closes the log before doing so. This is
2841 so that log files can straightforwardly be renamed and moved. However,
2842 there are a couple of unusual error situations where the daemon does write
2843 log entries, and I had neglected to close the log afterwards.
2845 30. The text of an SMTP error response that was received during a remote
2846 delivery was being truncated at 512 bytes. This is too short for some of
2847 the long messages that one sometimes sees. I've increased the limit to
2850 31. It is now possible to make retry rules that apply only when a message has a
2851 specific sender, in particular, an empty sender.
2853 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2854 it possible to be selective about when SMTP synchronization is enforced.
2856 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2858 32. Implemented hosts_connection_nolog.
2860 33. Added an ACL for QUIT.
2862 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2865 35. Added mailbox_size and mailbox_filecount to appendfile.
2867 36. Added control = no_multiline_responses to ACLs.
2869 37. There was a bug in the logic of the code that waits for the clock to tick
2870 in the case where the clock went backwards by a substantial amount such
2871 that the microsecond fraction of "now" was more than the microsecond
2872 fraction of "then" (but the whole seconds number was less).
2874 38. Added support for the libradius Radius client library this is found on
2875 FreeBSD (previously only the radiusclient library was supported).
2881 1. When certain lookups returned multiple values in the form name=value, the
2882 quoting of the values was not always being done properly. Specifically:
2883 (a) If the value started with a double quote, but contained no whitespace,
2885 (b) If the value contained whitespace other than a space character (i.e.
2886 tabs or newlines or carriage returns) it was not quoted.
2887 This fix has been applied to the mysql and pgsql lookups by writing a
2888 separate quoting function and calling it from the lookup code. The fix
2889 should probably also be applied to nisplus, ibase and oracle lookups, but
2890 since I cannot test any of those, I have not disturbed their existing code.
2892 2. A hit in the callout cache for a specific address caused a log line with no
2893 reason for rejecting RCPT. Now it says "Previous (cached) callout
2894 verification failure".
2896 3. There was an off-by-one bug in the queryprogram router. An over-long
2897 return line was truncated at 256 instead of 255 characters, thereby
2898 overflowing its buffer with the terminating zero. As well as fixing this, I
2899 have increased the buffer size to 1024 (and made a note to document this).
2901 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2902 when Exim is waiting for an SMTP response from a remote server, Exim
2903 restarts its select() call on the socket, thereby resetting its timeout.
2904 This is not a problem when such interrupts are rare. Somebody set up a cron
2905 job to run exiwhat every 2 minutes, which is less than the normal select()
2906 timeout (5 or 10 minutes). This meant that the select() timeout never
2907 kicked in because it was always reset. I have fixed this by comparing the
2908 time when an interrupt arrives with the time at the start of the first call
2909 to select(). If more time than the timeout has elapsed, the interrupt is
2910 treated as a timeout.
2912 5. Some internal re-factoring in preparation for the addition of Sieve
2913 extensions (by MH). In particular, the "personal" test is moved to a
2914 separate function, and given an option for scanning Cc: and Bcc: (which is
2915 not set for Exim filters).
2917 6. When Exim created an email address using the login of the caller as the
2918 local part (e.g. when creating a From: or Sender: header line), it was not
2919 quoting the local part when it contained special characters such as @.
2921 7. Installed new OpenBSD configuration files.
2923 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2924 try to make them clearer.
2926 9. Callout options, other than the timeout value, were being ignored when
2927 verifying sender addresses in header lines. For example, when using
2929 verify = header_sender/callout=no_cache
2931 the cache was (incorrectly) being used.
2933 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2934 only the Cygwin environment.
2936 11. When return_path_on_delivery was set as a log selector, if different remote
2937 addresses in the same message used different return paths and parallel
2938 remote delivery occurred, the wrong values would sometimes be logged.
2939 (Whenever a remote delivery process finished, the return path value from
2940 the most recently started remote delivery process was logged.)
2942 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2943 lines when AUTH and/or TLS are in use. This is the "received protocol"
2944 field. Exim used to use "asmtp" for authenticated SMTP, without any
2945 indication (in the protocol name) for TLS use. Now it follows the RFC and
2946 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2947 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2948 names appear in log lines as well as in Received: header lines.
2950 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2951 extensions, and comparison tests, and to fix some bugs.
2953 14. Changes to the "personal" filter test:
2955 (1) The test was buggy in that it was just doing the equivalent of
2956 "contains" tests on header lines. For example, if a user's address was
2957 anne@some.where, the "personal" test would incorrectly be true for
2959 To: susanne@some.where
2961 This test is now done by extracting each address from the header in turn,
2962 and checking the entire address. Other tests that are part of "personal"
2963 are now done using regular expressions (for example, to check local parts
2964 of addresses in From: header lines).
2966 (2) The list of non-personal local parts in From: addresses has been
2967 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2968 taken from the Sieve specification recommendations.
2970 (3) If the message contains any header line starting with "List-" it is
2971 treated as non-personal.
2973 (4) The test for "circular" in the Subject: header line has been removed
2974 because it now seems ill-conceived.
2976 15. Minor typos in src/EDITME comments corrected.
2978 16. Installed latest exipick from John Jetmore.
2980 17. If headers_add on a router specified a text string that was too long for
2981 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2982 of string_sprintf() is now avoided.
2984 18. $message_body_size was not set (it was always zero) when running the DATA
2985 ACL and the local_scan() function.
2987 19. For the "mail" command in an Exim filter, no default was being set for
2988 the once_repeat time, causing a random time value to be used if "once" was
2989 specified. (If the value happened to be <= 0, no repeat happened.) The
2990 default is now 0s, meaning "never repeat". The "vacation" command was OK
2991 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2992 (I found it when inspecting the code).
2994 20. There is now an overall timeout for performing a callout verification. It
2995 defaults to 4 times the callout timeout, which applies to individual SMTP
2996 commands during the callout. The overall timeout applies when there is more
2997 than one host that can be tried. The timeout is checked before trying the
2998 next host. This prevents very long delays if there are a large number of
2999 hosts and all are timing out (e.g. when the network connections are timing
3000 out). The value of the overall timeout can be changed by specifying an
3001 additional sub-option for "callout", called "maxwait". For example:
3003 verify = sender/callout=5s,maxwait=20s
3005 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3006 to the end before writing, but this should make it even safer).
3008 22. Exim was forgetting that it had advertised PIPELINING for the second and
3009 subsequent messages on an SMTP connection. It was also not resetting its
3010 memory on STARTTLS and an internal HELO.
3012 23. When Exim logs an SMTP synchronization error within a session, it now
3013 records whether PIPELINING has been advertised or not.
3015 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3016 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3017 rather than long int.
3019 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3021 26. Added the never_mail option to autoreply.
3027 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3028 crash if the getsockname() call failed; this can happen if a connection is
3029 closed very soon after it is established. The problem was simply in the
3030 order in which certain operations were done, causing Exim to try to write
3031 to the SMTP stream before it had set up the file descriptor. The bug has
3032 been fixed by making things happen in the correct order.
3038 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3039 before the connection was closed, thus losing the rejection response.
3041 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3042 some early Solaris releases, but causes trouble in current releases where
3043 socklen_t is defined.
3045 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3048 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3051 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3052 that releasing the top of it at the end releases what was used for sub-
3053 expansions (unless the block got too big). However, discard this block if
3054 the first thing is a variable or header, so that we can use its block when
3055 it is dynamic (useful for very large $message_headers, for example).
3057 6. Lookups now cache *every* query, not just the most recent. A new, separate
3058 store pool is used for this. It can be recovered when all lookup caches are
3059 flushed. Lookups now release memory at the end of their result strings.
3060 This has involved some general refactoring of the lookup sources.
3062 7. Some code has been added to the store_xxx() functions to reduce the amount
3063 of flapping under certain conditions.
3065 8. log_incoming_interface used to affect only the <= reception log lines. Now
3066 it causes the local interface and port to be added to several more SMTP log
3067 lines, for example "SMTP connection from", and rejection lines.
3069 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3071 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3073 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3074 could be overwritten at the end of the current message (or the start of a
3075 new message if it was set in a HELO ACL). The value is now preserved for
3076 the duration of the SMTP connection.
3078 12. If a transport had a headers_rewrite setting, and a matching header line
3079 contained an unqualified address, that address was qualified, even if it
3080 did not match any rewriting rules. The underlying bug was that the values
3081 of the flags that permit the existence of unqualified sender and recipient
3082 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3083 non-local messages, and by -bnq for local messages) were not being
3084 preserved with the message after it was received.
3086 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3087 "next input=" as part of the text comprising the host identity instead of
3088 the correct text. The code was using the same buffer for two different
3089 strings. However, depending on which order the printing function evaluated
3090 its arguments, the bug did not always show up. Under Linux, for example, my
3091 test suite worked just fine.
3093 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3094 doesn't work with some older versions of Perl. It has been changed to "my",
3095 which in any case is probably the better facility to use.
3097 15. A really picky compiler found some instances of statements for creating
3098 error messages that either had too many or two few arguments for the format
3101 16. The size of the buffer for calls to the DNS resolver has been increased
3102 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3103 for addresses that have a lot of PTR records. This alleviates a problem; it
3104 does not fully solve it.
3106 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3107 buffer now truncates the list and logs the incident, which is the same
3108 action as happens when Exim is looking up a host name and its aliases.
3109 Previously in this situation something unpredictable would happen;
3110 sometimes it was "internal error: store_reset failed".
3112 18. If a server dropped the connection unexpectedly when an Exim client was
3113 using GnuTLS and trying to read a response, the client delivery process
3114 crashed while trying to generate an error log message.
3116 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3117 string, for example:
3119 warn message = H1: something\nH2: something
3121 the text was added as a single header line from Exim's point of view
3122 though it ended up OK in the delivered message. However, searching for the
3123 second and subsequent header lines using $h_h2: did not work. This has been
3124 fixed. Similarly, if a system filter added multiple headers in this way,
3125 the routers could not see them.
3127 20. Expanded the error message when iplsearch is called with an invalid key to
3128 suggest using net-iplsearch in a host list.
3130 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3131 ACLs are no longer actually imposed (and a message to that effect is
3134 22. If a "gecos" field in a passwd entry contained escaped characters, in
3135 particular, if it contained a \" sequence, Exim got it wrong when building
3136 a From: or a Sender: header from that name. A second bug also caused
3137 incorrect handling when an unquoted " was present following a character
3138 that needed quoting.
3140 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3141 was not being matched caselessly.
3143 24. Arranged for all hyphens in the exim.8 source to be escaped with
3146 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3147 properly. Recipient callout cache records were still being keyed to include
3148 the sender, even when use_sender was set false. This led to far more
3149 callouts that were necessary. The sender is no longer included in the key
3150 when use_sender is false.
3152 26. Added "control = submission" modifier to ACLs.
3154 27. Added the ${base62d: operator to decode base 62 numbers.
3156 28. dnsdb lookups can now access SRV records.
3158 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3159 the configuration file.
3161 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3162 (-v) mode. This makes the output for a verbose queue run more intelligible.
3164 31. Added a use_postmaster feature to recipient callouts.
3166 32. Added the $body_zerocount variable, containing the number of binary zero
3167 bytes in the message body.
3169 33. The time of last modification of the "new" subdirectory is now used as the
3170 "mailbox time last read" when there is a quota error for a maildir
3173 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3175 35. Added +ignore_unknown as a special item in host lists.
3177 36. Code for decoding IPv6 addresses in host lists is now included, even if
3178 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3179 address was recognized as an IP address, but was then not correctly decoded
3180 into binary, causing unexpected and incorrect effects when compared with
3187 1. Very minor rewording of debugging text in manualroute to say "list of
3188 hosts" instead of "hostlist".
3190 2. If verify=header_syntax was set, and a header line with an unqualified
3191 address (no domain) and a large number of spaces between the end of the
3192 name and the colon was received, the reception process suffered a buffer
3193 overflow, and (when I tested it) crashed. This was caused by some obsolete
3194 code that should have been removed. The fix is to remove it!
3196 3. When running in the test harness, delay a bit after writing a bounce
3197 message to get a bit more predictability in the log output.
3199 4. Added a call to search_tidyup() just before forking a reception process. In
3200 theory, someone could use a lookup in the expansion of smtp_accept_max_
3201 per_host which, without the tidyup, could leave open a database connection.
3203 5. Added the variables $recipient_data and $sender_data which get set from a
3204 lookup success in an ACL "recipients" or "senders" condition, or a router
3205 "senders" option, similar to $domain_data and $local_part_data.
3207 6. Moved the writing of debug_print from before to after the "senders" test
3210 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3211 problems for message scanning, either using a data ACL, or using
3212 local_scan() because the Received: header was not generated till after they
3213 were called (in order to set the time as the time of reception completion).
3214 I have revised the way this works. The header is now generated after the
3215 body is received, but before the ACL or local_scan() are called. After they
3216 are run, the timestamp in the header is updated.
3222 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3223 before starting a queue runner without re-exec. This happened only when
3224 deliver_drop_privilege was set or when the Exim user was set to root. The
3225 effect of the bug was that timeouts during subsequent deliveries caused
3226 crashes instead of being properly handled. The handler is now left at its
3227 default (and expected) setting.
3229 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3230 message, again when deliver_drop_privilege is set or Exim is run as root.
3231 The bug described in (1) was not present in this case, but the tidying up
3232 of the other signals was missing. I have made the two cases consistent.
3234 3. The ignore_target_hosts setting on a manualroute router was being ignored
3235 for hosts that were looked up using the /MX notation.
3237 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3240 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3241 operated on the sender address. After changing the $sender_address to <>
3242 for the sender address verify, Exim was re-instated it as the original
3243 (before rewriting) address, but remembering that it had rewritten it, so it
3244 wasn't rewriting it again. This bug also had the effect of breaking the
3245 sender address verification caching when the sender address was rewritten.
3247 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3248 This has been changed so that if the ip literal address matches
3249 ignore_target_hosts, the router declines.
3251 7. Added expansion conditions match_domain, match_address, and match_local_
3252 part (NOT match_host).
3254 8. The placeholder for the Received: header didn't have a length field set.
3256 9. Added code to Exim itself and to exim_lock to test for a specific race
3257 condition that could lead to file corruption when using MBX delivery. The
3258 issue is with the lockfile that is created in /tmp. If this file is removed
3259 after a process has opened it but before that process has acquired a lock,
3260 there is the potential for a second process to recreate the file and also
3261 acquire a lock. This could lead to two Exim processes writing to the file
3262 at the same time. The added code performs the same test as UW imapd; it
3263 checks after acquiring the lock that its file descriptor still refers to
3264 the same named file.
3266 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3267 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3270 11. Added the smtp_active_hostname option. If used, this will typically be made
3271 to depend on the incoming interface address. Because $interface_address is
3272 not set up until the daemon has forked a reception process, error responses
3273 that can happen earlier (such as "too many connections") no longer contain
3276 12. If an expansion in a condition on a "warn" statement fails because a lookup
3277 defers, the "warn" statement is abandoned, and the next ACL statement is
3278 processed. Previously this caused the whole ACL to be aborted.
3280 13. Added the iplsearch lookup type.
3282 14. Added ident_timeout as a log selector.
3284 15. Added tls_certificate_verified as a log selector.
3286 16. Added a global option tls_require_ciphers (compare the smtp transport
3287 option of the same name). This controls incoming TLS connections.
3289 17. I finally figured out how to make tls_require_ciphers do a similar thing
3290 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3291 before starting the TLS session.
3293 18. Tabs are now shown as \t in -bP output.
3295 19. If the log selector return_path_on_delivery was set, Exim crashed when
3296 bouncing a message because it had too many Received: header lines.
3298 20. If two routers both had headers_remove settings, and the first one included
3299 a superfluous trailing colon, the final name in the first list and the
3300 first name in the second list were incorrectly joined into one item (with a
3301 colon in the middle).
3307 1. Added -C and -D options to the exinext utility, mainly to make it easier
3308 to include in the automated testing, but these could be helpful when
3309 multiple configurations are in use.
3311 2. The exinext utility was not formatting the output nicely when there was
3312 an alternate port involved in the retry record key, nor when there was a
3313 message id as well (for retries that were specific to a specific message
3314 and a specific host). It was also confused by IPv6 addresses, because of
3315 the additional colons they contain. I have fixed the IPv4 problem, and
3316 patched it up to do a reasonable job for IPv6.
3318 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3319 delivery, the log line now contains "pipelined" if PIPELINING was used.
3321 4. An SMTP transport process used to panic and die if the bind() call to set
3322 an explicit outgoing interface failed. This has been changed; it is now
3323 treated in the same way as a connect() failure.
3325 5. A reference to $sender_host_name in the part of a conditional expansion
3326 that was being skipped was still causing a DNS lookup. This no longer
3329 6. The def: expansion condition was not recognizing references to header lines
3330 that used bh_ and bheader_.
3332 7. Added the _cache feature to named lists.
3334 8. The code for checking quota_filecount in the appendfile transport was
3335 allowing one more file than it should have been.
3337 9. For compatibility with Sendmail, the command line option
3345 and sets the incoming protocol and host name (for trusted callers). The
3346 host name and its colon can be omitted when only the protocol is to be set.
3347 Note the Exim already has two private options, -pd and -ps, that refer to
3348 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3349 "s", but I don't think that's a major issue.
3351 10. A number of refactoring changes to the code, none of which should affect
3354 (a) The number of logging options was getting close to filling up the
3355 32-bit word that was used as a bit map. I have split them into two classes:
3356 those that are passed in the argument to log_write(), and those that are
3357 only ever tested independently outside of that function. These are now in
3358 separate 32-bit words, so there is plenty of room for expansion again.
3359 There is no change in the user interface or the logging behaviour.
3361 (b) When building, for example, log lines, the code previously used a
3362 macro that called string_cat() twice, in order to add two strings. This is
3363 not really sufficiently general. Furthermore, there was one instance where
3364 it was actually wrong because one of the argument was used twice, and in
3365 one call a function was used. (As it happened, calling the function twice
3366 did not affect the overall behaviour.) The macro has been replaced by a
3367 function that can join an arbitrary number of extra strings onto a growing
3370 (c) The code for expansion conditions now uses a table and a binary chop
3371 instead of a serial search (which was left over from when there were very
3372 few conditions). Also, it now recognizes conditions like "pam" even when
3373 the relevant support is not compiled in: a suitably worded error message is
3374 given if an attempt is made to use such a condition.
3376 11. Added ${time_interval:xxxxx}.
3378 12. A bug was causing one of the ddress fields not to be passed back correctly
3379 from remote delivery subprocesses. The field in question was not being
3380 subsequently used, so this caused to problems in practice.
3382 13. Added new log selectors queue_time and deliver_time.
3384 14. Might have fixed a bug in maildirsizefile handling that threw up
3385 "unexpected character" debug warnings, and recalculated the data
3386 unnecessarily. In any case, I expanded the warning message to give more
3389 15. Added the message "Restricted characters in address" to the statements in
3390 the default ACL that block characters like @ and % in local parts.
3392 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3393 Three changes have been made:
3395 (a) There was a serious bug; a negative response to MAIL caused the whole
3396 recipient domain to be cached as invalid, thereby blocking all messages
3397 to all local parts at the same domain, from all senders. This bug has
3398 been fixed. The domain is no longer cached after a negative response to
3399 MAIL if the sender used is not empty.
3401 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3404 (c) A new callout option, "use_sender" has been added for people who want
3405 the modified behaviour.
3411 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3414 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3415 indeed breaks things for older releases.
3417 3. Added additional logging to the case where there is a problem reading data
3418 from a filter that is running in a subprocess using a pipe, in order to
3419 try to track down a specific problem.
3421 4. Testing facility fudge: when running in the test harness and attempting
3422 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3423 getting "No route to host". Convert this to a timeout.
3425 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3428 6. Some OS don't have socklen_t but use size_t instead. This affects the
3429 fifth argument of getsockopt() amongst other things. This is now
3430 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3431 can be set for individual OS. I have set it for SunOS5, OSF1, and
3432 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3433 some earlier ones do not.
3435 7. Change 4.30/15 was not doing the test caselessly.
3437 8. The standard form for an IPv6 address literal was being rejected by address
3438 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3439 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3440 this, as well as the form without the "IPv6" on the front (but only when
3441 address literals are enabled, of course).
3443 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3445 10. Exim crashed if a message with an empty sender address specified by -f
3446 encountered a router with an errors_to setting. This could be provoked only
3447 by a command such as
3451 where an empty string was supplied; "<>" did not hit this bug.
3453 11. Installed PCRE release 4.5.
3455 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3456 remained set. It is now erased.
3458 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3459 times from message ids (which are base 36 rather than the normal 62).
3461 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3462 were being counted as actual protocol errors, and logged if the log
3463 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3464 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3465 and DATA following a set of rejected RCPTs do not count as protocol errors.
3466 In other words, Exim assumes they were pipelined, though this may not
3467 actually be the case. Of course, in all cases the client gets an
3468 appropriate error code.
3470 15. If a lookup fails in an ACL condition, a message about the failure may
3471 be available; it is used if testing the ACL cannot continue, because most
3472 such messages specify what the cause of the deferral is. However, some
3473 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3474 that caused an old message to be retained and used if a later statement
3475 caused a defer, replacing the real cause of the deferral.
3477 16. If an IP address had so many PTR records that the DNS lookup buffer
3478 was not large enough to hold them, Exim could crash while trying to process
3479 the truncated data. It now detects and logs this case.
3481 17. Further to 4.21/58, another change has been made: if (and only if) the
3482 first line of a message (the first header line) ends with CRLF, a bare LF
3483 in a subsequent header line has a space inserted after it, so as not to
3484 terminate the header.
3486 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3487 unnecessarily, used atoi() instead of strtol(), and didn't check the
3488 termination when getting file sizes from file names by regex.
3490 19. Completely re-implemented the support for maildirsize files, in the light
3491 of a number of problems with the previous contributed implementation
3492 (4.30/29). In particular:
3494 . If the quota is zero, the maildirsize file is maintained, but no quota is
3497 . If the maildir directory does not exist, it is created before any attempt
3498 to write a maildirsize file.
3500 . The quota value in the file is just a cache; if the quota is changed in
3501 the transport, the new value overrides.
3503 . A regular expression is available for excluding directories from the
3506 20. The autoreply transport checks the characters in options that define the
3507 message's headers; it allows continued headers, but it was checking with
3508 isspace() after an embedded newline instead of explicitly looking for a
3511 21. If all the "regular" hosts to which an address was routed had passed their
3512 expiry times, and had not reached their retry times, the address was
3513 bounced, even if fallback hosts were defined. Now Exim should go on to try
3516 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3517 equivalent code in the SMTP transport. Some hosts send humungous responses
3518 to HELO/EHLO, more than 1024 it seems.
3520 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3521 gives compiler warnings in some environments. I've now done it "properly",
3524 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3525 (because of problems with the built-in one) was declared to return uschar *
3526 instead of char *, causing compiler failure.
3528 25. Fixed a file descriptor leak when processing alias/forward files.
3530 26. Fixed a minor format string issue in dbfn.c.
3532 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3534 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3535 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3538 29. When a sender address is verified, it is cached, to save repeating the test
3539 when there is more than one recipient in a message. However, when the
3540 verification involves a callout, it is possible for different callout
3541 options to be set for different recipients. It is too complicated to keep
3542 track of this in the cache, so now Exim always runs a verification when a
3543 callout is required, relying on the callout cache for the optimization.
3544 The overhead is duplication of the address routing, but this should not be
3547 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3548 to be verified with callout=postmaster, and the main callout worked but the
3549 postmaster check failed, the verification correctly failed. However, if a
3550 subsequent RCPT command asked for sender verification *without* the
3551 postmaster check, incorrect caching caused this verification also to fail,
3554 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3555 it was not caching the DNS options (qualify_single, search_parents) that
3556 were used when the lookup failed. A subsequent lookup with different
3557 options therefore always gave the same answer, though there were cases
3558 where it should not have. (Example: a "domains = !$mx_any" option on a
3559 dnslookup router: the "domains" option is always processed without any
3560 widening, but the router might have qualify_single set.) Now Exim uses the
3561 cached value only when the same options are set.
3563 32. Added John Jetmore's "exipick" utility to the distribution.
3565 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3566 than a timeout (e.g. a certificate is required, and is not provided), an
3567 Exim server now closes the connection immediately. Previously it waited for
3568 the client to close - but if the client is SSL, it seems that they each
3569 wait for each other, leading to a delay before one of them times out.
3571 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3572 maintained 0.8.x compatibility because I don't think many are using it, and
3573 it is clearly obsolete.
3575 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3578 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3579 client certificate was expired. A simple patch fixes this, though I don't
3580 understand the full logic of why the verify callback is called multiple
3583 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3584 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3585 which causes problems with some clients (such as the Certicom SSL Plus
3586 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3587 disables the coutermeasure allowing Eudora to connect."
3589 38. Exim was not checking that a write() to a log file succeeded. This could
3590 lead to Bad Things if a log got too big, in particular if it hit a file
3591 size limit. Exim now panics and dies if it cannot write to a log file, just
3592 as it does if it cannot open a log file.
3594 39. Modified OS/Makefile-Linux so that it now contains
3596 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3598 The two -D definitions ensure that Exim is compiled with large file
3599 support, which makes it possible to handle log files that are bigger than
3602 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3603 instance) a domain was checked against a named list that involved a lookup,
3604 causing $domain_data to be set, then another domain was checked against the
3605 same list, then the first domain was re-checked, the value of $domain_data
3606 after the final check could be wrong. In particular, if the second check
3607 failed, it could be set empty. This bug probably also applied to
3610 41. The strip_trailing_dot option was not being applied to the address given
3611 with the -f command-line option.
3613 42. The code for reading a message's header from the spool was incrementing
3614 $received_count, but never initializing it. This meant that the value was
3615 incorrect (doubled) while delivering a message in the same process in which
3616 it was received. In the most common configuration of Exim, this never
3617 happens - a fresh exec is done - but it can happen when
3618 deliver_drop_privilege is set.
3620 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3621 it now includes up to 150 characters of the unexpected data in the log
3624 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3625 and building data strings. The size of both of these buffers was 10 000
3626 bytes - far larger than anybody would *ever* want, thought I. Needless to
3627 say, somebody hit the limit. I have increased the maximum line length to
3628 20 000 and the maximum data length of concatenated lines to 100 000. I have
3629 also fixed two bugs, because there was no checking on these buffers. Tsk,
3630 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3631 buffer is too small.
3633 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3634 lsearch lookups. Now it does.
3636 46. When parsing a route_list item in a manualroute router, a fixed-length
3637 buffer was used for the list of hosts. I made this 1024 bytes long,
3638 thinking that nobody would ever have a list of hosts that long. Wrong.
3639 Somebody had a whole pile of complicated expansion conditions, and the
3640 string was silently truncated, leading to an expansion error. It turns out
3641 that it is easier to change to an unlimited length (owing to other changes
3642 that have happened since this code was originally written) than to build
3643 structure for giving a limitation error. The length of the item that
3644 expands into the list of hosts is now unlimited.
3646 47. The lsearch lookup could not handle data where the length of text line was
3647 more than 4095 characters. Such lines were truncated, leading to shortened
3648 data being returned. It should now handle lines of any length.
3650 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3651 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3654 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3655 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3656 no longer happens in comment lines. A list of replacements is now placed
3657 at the head of all of the source files, except those whose only change is
3658 to replace PERL_COMMAND in the very first #! line.
3660 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3661 messages on the queue, with a bottom-up merge sort, using code contributed
3662 by Michael Haardt. This should make operations like -bp somewhat faster on
3663 large queues. It won't affect queue runners, except when queue_run_in_order
3666 51. Installed eximstats 1.31 in the distribution.
3668 52. Added support for SRV lookups to the dnslookup router.
3670 53. If an ACL referred to $message_body or $message_body_end, the value was not
3671 reset for any messages that followed in the same SMTP session.
3673 54. The store-handling optimization for building very long strings was not
3674 differentiating between the different store pools. I don't think this
3675 actually made any difference in practice, but I've tidied it.
3677 55. While running the routers to verify a sender address, $sender_address
3678 was still set to the sender address. This is wrong, because when routing to
3679 send a bounce to the sender, it would be empty. Therefore, I have changed
3680 it so that, while verifying a sender address, $sender_address is set to <>.
3681 (There is no change to what happens when verifying a recipient address.)
3683 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3684 target A or AAAA records (if not already returned) without resetting the
3685 qualify_single or search_parents options of the DNS resolver. These are
3686 inappropriate in this case because the targets of MX and SRV records must
3687 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3688 target that, when qualified, matched something in the local domain. These
3689 two options are now turned off when doing these lookups.
3691 57. It seems that at least some releases of Reiserfs (which does not have the
3692 concept of a fixed number of inodes) returns zero and not -1 for the
3693 number of available inodes. This interacted badly with check_spool_inodes,
3694 which assumed that -1 was the "no such thing" setting. What I have done is
3695 to check that the total number of inodes is greater than zero before doing
3696 the test of how many are available.
3698 58. When a "warn" ACL statement has a log_message modifier, the message is
3699 remembered, and not repeated. This is to avoid a lot of repetition when a
3700 message has many recipients that cause the same warning to be written.
3701 Howewer, Exim was preserving the list of already written lines for an
3702 entire SMTP session, which doesn't seem right. The memory is now reset if a
3703 new message is started.
3705 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3706 debugging output unless log_rewrite was also set.
3708 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3709 of (int)(handle) when we know that handle contains (void *)(-1).
3711 61. The Exim daemon panic-logs an error return when it closes the incoming
3712 connection. However "connection reset by peer" seems to be common, and
3713 isn't really an error worthy of noting specially, so that particular error
3716 62. When Exim is trying to find all the local interfaces, it used to panic and
3717 die if the ioctl to get the interface flags failed. However, it seems that
3718 on at least one OS (Solaris 9) it is possible to have an interface that is
3719 included in the list of interfaces, but for which you get a failure error
3720 for this call. This happens when the interface is not "plumbed" into a
3721 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3722 failure of the "get flags" call assumes that the interface is down.
3724 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3725 makes life easier for people who are doing arithmetic on fields extracted
3726 from dates, where you often get leading zeros that should not be
3727 interpreted as octal.
3729 64. Added qualify_domain to the redirect router, to override the global
3732 65. If a pathologically long header line contained very many addresses (the
3733 report of this problem mentioned 10 000) and each of them was rewritten,
3734 Exim could use up a very large amount of memory. (It kept on making new
3735 copies of the header line as it rewrote, and never released the old ones.)
3736 At the expense of a bit more processing, the header rewriting function has
3737 been changed so that it no longer eats memory in this way.
3739 66. The generation of the Received: header has been moved from the time that a
3740 message starts to be received, to the time that it finishes. The timestamp
3741 in the Received: header should now be very close to that of the <= log
3742 line. There are two side-effects of this change:
3744 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3745 logged header lines no longer include the local Received: line, because
3746 it has not yet been created. The same applies to a copy of the message
3747 that is returned to a non-SMTP sender when a message is rejected.
3749 (b) When a filter file is tested using -bf, no additional Received: header
3750 is added to the test message. After some thought, I decided that this
3753 This change does not affect the value of $received_for. It is still set
3754 after address rewriting, but before local_scan() is called.
3756 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3758 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3759 gave an unhelpful panic error message, and a defer error. I have managed to
3760 change this behaviour so that it now rejects any supplied certificate,
3761 which seems right, as the list of acceptable certificates is empty.
3763 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3764 gave an unhelpful defer error. I have not managed to make this reject any
3765 supplied certificates, but the error message it gives is "no certificate
3766 supplied", which is not helpful.
3768 70. exigrep's output now also includes lines that are not associated with any
3769 message, but which match the given pattern. Implemented by a patch from
3770 Martin Sluka, which also tidied up the Perl a bit.
3772 71. Recipient callout verification, like sender verification, was using <> in
3773 the MAIL FROM command. This isn't really the right thing, since the actual
3774 sender may affect whether the remote host accepts the recipient or not. I
3775 have changed it to use the actual sender in the callout; this means that
3776 the cache record is now keyed on a recipient/sender pair, not just the
3777 recipient address. There doesn't seem to be a real danger of callout loops,
3778 since a callout by the remote host to check the sender would use <>.
3779 [SEE ABOVE: changed after hitting problems.]
3781 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3782 temporary errors. However, in the case of such a code being given after
3783 the end of a data transmission (i.e. after ".") Exim was failing to write
3784 a retry record for the message. (Yes, there was some broken host that was
3785 actually sending 8xx at this point.)
3787 73. An unknown lookup type in a host list could cause Exim to panic-die when
3788 the list was checked. (An example that provoked this was putting <; in the
3789 middle of a list instead of at the start.) If this happened during a DATA
3790 ACL check, a -D file could be left lying around. This kind of configuration
3791 error no longer causes Exim to die; instead it causes a defer errror. The
3792 incident is still logged to the main and panic logs.
3794 74. Buglet left over from Exim 3 conversion. The message "too many messages
3795 in one connection" was written to the rejectlog but not the mainlog, except
3796 when address rewriting (yes!) was being logged.
3798 75. Added write_rejectlog option.
3800 76. When a system filter was run not as root (that is, when system_filter_user
3801 was set), the values of the $n variables were not being returned to the
3802 main process; thus, they were not subsequently available in the $sn
3805 77. Added +return_path_on_delivery log selector.
3807 78. A connection timeout was being treated differently from recipients deferred
3808 when testing hosts_max_try with a message that was older than the host's
3809 retry timeout. (The host should not be counted, thus allowing all hosts to
3810 be tried at least once before bouncing.) This may have been the cause of an
3811 occasionally reported bug whereby a message would remain on the queue
3812 longer than the retry timeout, but would be bounced if a delivery was
3813 forced. I say "may" because I never totally pinned down the problem;
3814 setting up timeout/retry tests is difficult. See also the next item.
3816 79. The ultimate address timeout was not being applied to errors that involved
3817 a combination of host plus message (for example, a timeout on a MAIL
3818 command). When an address resolved to a number of possible hosts, and they
3819 were not all tried for each delivery (e.g. because of hosts_max_try), a
3820 message could remain on the queue longer than the retry timeout.
3822 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3825 81. Fixed an obscure SMTP outgoing bug which required at least the following
3826 conditions: (a) there was another message waiting for the same server;
3827 (b) the server returned 5xx to all RCPT commands in the first message so
3828 that the message was not completed; (c) the server dropped the connection
3829 or gave a negative response to the RSET that Exim sends to abort the
3830 transaction. The observed case was a dropped connection after DATA that had
3831 been sent in pipelining mode. That is, the server had advertised PIPELINING
3832 but was not implementing it correctly. The effect of the bug was incorrect
3833 behaviour, such as trying another host, and this could lead to a crash.
3839 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3840 and daemon.c were passed as pointers to ints; they should have been
3841 pointers to socklen_t variables (which are typically unsigned ints).
3843 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3846 3. Fixed a really odd bug that affected only the testing scheme; patching a
3847 certain fixed string in the binary changed the value of another string that
3848 happened to be identical to the end of the original first string.
3850 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3851 name", it returns that address as the IP address. On some operating
3852 systems (e.g. Solaris), it also passes back the IP address string as the
3853 "host name". However, on others (e.g. Linux), it passes back an empty
3854 string. Exim wasn't checking for this, and was changing the host name to an
3855 empty string, assuming it had been canonicized.
3857 5. Although rare, it is permitted to have more than one PTR record for a given
3858 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3859 all the names associated with an address, because they do in Solaris.
3860 However, it seems that they do not in Linux for data that comes from the
3861 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3862 I found this out when I moved to a new Linux workstation and tried to run
3863 the Exim test suite.
3865 To get round this problem I have changed the code so that it now does its
3866 own call to the DNS to look up PTR records when searching for a host name.
3867 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3868 addresses that are only in /etc/hosts are still found.
3870 This behaviour is, however, controlled by an option called host_lookup_
3871 order, which defaults to "bydns:byaddr". If people want to use the other
3872 order, or indeed, just use one or the other means of lookup, they can
3873 specify it in this variable.
3875 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3876 some operating systems, this comes back from gethostbyaddr() as an empty
3877 string, and this is what Exim used to test for. However, it seems that in
3878 other systems, "." is yielded. Exim now tests for this case too.
3880 7. The values of check_spool_space and check_log_space are now held internally
3881 as a number of kilobytes instead of an absolute number of bytes. If a
3882 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3883 kilobyte. This means that much larger values can be stored.
3885 8. Exim monitor: an attempt to get the action menu when not actually pointing
3886 at a message produces an empty menu entitled "No message selected". This
3887 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3888 no entries in it ("Shell widget menu has zero width and/or height"). So I
3889 have added a single, blank menu entry in this case.
3891 9. Added ${quote_local_part.
3893 10. MIME decoding is now applied to the contents of Subject: header lines when
3896 11. Now that a reference to $sender_host_address automatically causes a reverse
3897 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3898 host lookup before query-style lookups in lists that might use this
3899 variable. This has therefore been abolished, and the "net-" prefix is no
3900 longer necessary for query-style lookups.
3902 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3903 have been a typo for LFLAGS, so it has been changed.
3905 13. The install script calls Exim with "-C /dev/null" in order to find the
3906 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3907 to be output. Howeve, since Exim outputs its version number before the
3908 error, it didn't break the script. It just looked ugly. I fixed this by
3909 always allowing "-C /dev/null" if the caller is root.
3911 14. Ignore overlarge ACL variable number when reading spool file - insurance
3912 against a later release with more variables having written the file.
3914 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3915 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3916 this, as well as the form without the "IPv6" on the front.
3918 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3919 OS/Makefile-Darwin file.
3921 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3922 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3924 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3925 variable. However, some OS use ints for inodes. I've added cast to long int
3926 to get rid of the compiler warning.
3928 19. I had forgotten to lock out "/../" in configuration file names when
3929 ALT_CONFIG_PREFIX was set.
3931 20. Routers used for verification do not need to specify transports. However,
3932 if such a router generated a host list, and callout was configured, Exim
3933 crashed, because it could not find a port number from the (non-existent)
3934 transport. It now assumes port 25 in this circumstance.
3936 21. Added the -t option to exigrep.
3938 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3939 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3940 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3941 them is set, LOOKUP_LSEARCH is forced.
3943 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3945 24. Added sender and host information to the "rejected by local_scan()" log
3946 line; previously there was no indication of these.
3948 25. Added .include_if_exists.
3950 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3951 Linux. It turns out that not all file systems support this. Apparently some
3952 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3953 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3954 not supported on the file descriptor, is now ignored when Exim is trying to
3955 sync a directory. This applies only to Linux.
3957 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3959 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3960 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3963 29. Added support for maildirsize files from supplied patch (modified a bit).
3965 30. The use of :fail: followed by an empty string could lead Exim to respond to
3966 sender verification failures with (e.g.):
3968 550 Verification failed for <xxx>
3969 550 Sender verify failed
3971 where the first response line was missing the '-' that indicates it is not
3972 the final line of the response.
3974 31. The loop for finding the name of the user that called Exim had a hardwired
3975 limit of 10; it now uses the value of finduser_retries, which is used for
3976 all other user lookups.
3978 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3981 33. Exim was neglecting to zero errno before one call of strtol() when
3982 expanding a string and expecting an integer value. On some systems this
3983 resulted in spurious "integer overflow" errors. Also, it was casting the
3984 result into an int without checking.
3986 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3987 did not work. The code looks as if it has *never* worked, though it appears
3988 to have been documented since at least releast 1.62. I have made it work.
3990 35. The "timeout_DNS" error in retry rules, also documented since at least
3991 1.62, also never worked. As it isn't clear exactly what this means, and
3992 clearly it isn't a major issue, I have abolished the feature by treating it
3993 as "timeout", and writing a warning to the main and panic logs.
3995 36. The display of retry rules for -brt wasn't always showing the error code
3998 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3999 timeout_connect_A, timeout_connect_MX.
4001 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4002 to the empty sender.
4004 39. The daemon was not analysing the content of -oX till after it had closed
4005 stderr and disconnected from the controlling terminal. This meant that any
4006 syntax errors were only noted on the panic log, and the return code from
4007 the command was 0. By re-arranging the code a little, I've made the
4008 decoding happen first, so such errors now appear on stderr, and the return
4009 code is 1. However, the actual setting up of the sockets still happens in
4010 the disconnected process, so errors there are still only recorded on the
4013 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4014 connections (as happens on some IP stacks) was logged at start up time as
4015 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4016 it from "IPv6 and IPv4", which means that two separate sockets are being
4019 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4020 says whether AF_INET or AF_INET6 was passed as an argument.
4022 42. Exiwhat output was messed up when time zones were included in log
4025 43. Exiwhat now gives more information about the daemon's listening ports,
4026 and whether -tls-on-connect was used.
4028 44. The "port" option of the smtp transport is now expanded.
4030 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4031 silently ignored. Now an error message is written to the main and panic
4034 46. There's a new ACL modifier called "logwrite" which writes to a log file
4035 as soon as it is encountered.
4037 47. Added $local_user_uid and $local_user_gid at routing time.
4039 48. Exim crashed when trying to verify a sender address that was being
4042 49. Exim was recognizing only a space character after ".include". It now also
4043 recognizes a tab character.
4045 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4046 extracting the relevant information from the specification. The man page no
4047 longer contains scrambled data for the -d option, and I've added a section
4048 at the front about calling Exim under different names.
4050 51. Added "extra_headers" argument to the "mail" command in filter files.
4052 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4055 53. Installed eximstats 1.29.
4057 54. Added transport_filter_timeout as a generic transport option.
4059 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4060 Cc: header lines. This was required by RFC 822, but it not required by RFC
4063 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4064 incoming messages that did not have them. Now it does so only if the
4065 message originates locally, that is, if there is no associated remote host
4066 address. When Resent- header lines are present, this applies to the Resent-
4067 lines rather than the non-Resent- lines.
4069 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4070 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4072 58. Messages for configuration errors now include the name of the main
4073 configuration file - useful now that there may be more than one file in a
4074 list (.included file names were always shown).
4076 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4077 for those rare installations that do not start the daemon as root or run it
4078 setuid root. I've cut out the call to initgroups() if the daemon is not
4081 60. The Exim user and group can now be bound into the binary as text strings
4082 that are looked up at the start of Exim's processing.
4084 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4086 62. Added $mailstore_basename variable.
4088 63. Installed patch to sieve.c from Michael Haardt.
4090 64. When Exim failed to open the panic log after failing to open the main log,
4091 the original message it was trying to log was written to stderr and debug
4092 output, but if they were not available (the usual case in production), it
4093 was lost. Now it is written to syslog before the two lines that record the
4094 failures to open the logs.
4096 65. Users' Exim filters run in subprocesses under the user's uid. It is
4097 possible for a "deliver" command or an alias in a "personal" command to
4098 provoke an address rewrite. If logging of address rewriting is configured,
4099 this fails because the process is not running as root or exim. There may be
4100 a better way of dealing with this, but for the moment (because 4.30 needs
4101 to be released), I have disabled address rewrite logging when running a
4102 filter in a non-root, non-exim process.
4108 1. The buildconfig auxiliary program wasn't quoting the value set for
4109 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4110 not defined. This bug was masked in 4.22 by the effect that was fixed in
4113 2. Some messages that were rejected after a message id was allocated were
4114 shown as "incomplete" by exigrep. It no longer does this for messages that
4115 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4117 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4118 have allow_domain_literals set, the ID did not get logged in the <= line.
4119 Domain literals are now always recognized in Message-ID: header lines.
4121 4. The first argument for a ${extract expansion item is the key name or field
4122 number. Leading and trailing spaces in this item were not being ignored,
4123 causing some misleading effects.
4125 5. When deliver_drop_privilege was set, single queue runner processes started
4126 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4127 same command in the process it spins off) were not dropping privilege.
4129 6. When the daemon running as "exim" started a queue runner, it always
4130 re-executed Exim in the spun-off process. This is a waste of effort when
4131 deliver_drop_privilege is set. The new process now just calls the
4132 queue-runner function directly.
4138 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4141 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4142 ignored. Though the use of -oP was forcing the writing of a pid file, it
4143 was always written to the default place.
4145 3. If the message "no IP address found for host xxxx" is generated during
4146 incoming verification, it is now followed by identification of the incoming
4147 connection (so you can more easily find what provoked it).
4149 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4151 5. Added some features to "harden" Exim a bit more against certain attacks:
4153 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4154 be put in Local/Makefile. This is like the never_users runtime option,
4155 but it cannot be overridden. The default setting is "root".
4157 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4158 prefix string with which any file named in a -C command line option
4161 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4162 is retained for -C and -D only if the caller of Exim is root. Without
4163 it, the exim user may also use -C and -D and retain privilege.
4165 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4166 command line option is disabled.
4168 6. Macro names set by the -D option must start with an upper case letter, just
4169 like macro names defined in the configuration file.
4171 7. Added "dereference=" facility to LDAP.
4173 8. Two instances of the typo "uknown" in the source files are fixed.
4175 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4176 the Configure-Makefile script screwed up while processing it.
4178 10. Incorporated PCRE 4.4.
4180 11. The SMTP synchronization check was not operating right at the start of an
4181 SMTP session. For example, it could not catch a HELO sent before the client
4182 waited for the greeting. There is now a check for outstanding input at the
4183 point when the greeting is written. Because of the duplex, asynchronous
4184 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4185 way, but not yet received, when the check is performed.
4187 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4188 on TCP/IP sockets, because this apparently causes some broken clients to
4191 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4192 unchanged) from the Cygwin maintainer.
4194 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4195 was supported instead of testing for mbx. Effectively a typo.
4197 15. The spa authenticator server code was not checking that the input it
4198 received was valid base64.
4200 16. The debug output line for the "set" modifier in ACLs was not showing the
4201 name of the variable that was being set.
4203 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4205 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4206 lookup on its own. Something else was needed to trigger the lookup. For
4207 example, a match in host_lookup or the need for a host name in a host list.
4208 Now, if $sender_host_name is referenced and the host name has not yet been
4209 looked up, a lookup is performed. If the lookup fails, the variable remains
4210 empty, and $host_lookup_failed is set to "1".
4212 19. Added "eqi" as a case-independent comparison operator.
4214 20. The saslauthd authentication condition could segfault if neither service
4215 nor realm was specified.
4217 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4218 error message that was logged was misleading, and incoming SMTP
4219 connections were dropped. The message is now more accurate, and temporary
4220 errors are given to SMTP connections.
4222 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4223 (except RSET) with a 421 error, until QUIT is received. However, it was
4224 failing to send a response to QUIT.
4226 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4227 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4228 ACL is now run after the helo_try_verify_hosts code.
4230 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4231 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4232 case-independent, so other case variants are also recognized). Apparently
4233 some systems use these upper case variants.
4235 25. If more than two messages were waiting for the same host, and a transport
4236 filter was specified for the transport, Exim sent two messages over the
4237 same TCP/IP connection, and then failed with "socket operation on non-
4238 socket" when it tried to send the third.
4240 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4242 27. The extern definition of crypt16() in expand.c was not being excluded when
4243 the OS had its own crypt16() function.
4245 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4246 as a preferred synonym for return_size_limit, both as an option and as an
4249 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4251 30. Changed the default configuration ACL to relax the local part checking rule
4252 for addresses that are not in any local domains. For these addresses,
4253 slashes and pipe symbols are allowed within local parts, but the sequence
4254 /../ is explicitly forbidden.
4256 31. SPA server authentication was not clearing the challenge buffer before
4259 32. log_message in a "warn" ACL statement was writing to the reject log as
4260 well as to the main log, which contradicts the documentation and doesn't
4261 seem right (because no rejection is happening). So I have stopped it.
4263 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4264 However, I am unable to do any testing of this.
4266 34. Fixed an infelicity in the appendfile transport. When checking directories
4267 for a mailbox, to see if any needed to be created, it was accidentally
4268 using path names with one or more superfluous leading slashes; tracing
4269 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4271 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4272 contents are added to the log line that is written for every discarded
4273 recipient. (Previously a log_message setting was ignored.)
4275 36. The ${quote: operator now quotes the string if it is empty.
4277 37. The install script runs exim in order to find its version number. If for
4278 some reason other than non-existence or emptiness, which it checks, it
4279 could not run './exim', it was installing it with an empty version number,
4280 i.e. as "exim-". This error state is now caught, and the installation is
4283 38. An argument was missing from the function that creates an error message
4284 when Exim fails to connect to the socket for saslauthd authentication.
4285 This could cause Exim to crash, or give a corrupted message.
4287 39. Added isip, isip4, and isip6 to ${if conditions.
4289 40. The ACL variables $acl_xx are now saved with the message, and can be
4290 accessed later in routers, transports, and filters.
4292 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4293 strings in the file are not string-expanded.
4295 42. If a MAIL command specified a SIZE value that was too large to fit into an
4296 int variable, the check against message_size_limit failed. Such values are
4297 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4298 day this will have to be increased, but I don't think I want to be around
4299 when emails are that large.
4306 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4307 iconv() is not standard in FreeBSD.
4309 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4310 IPv6 enabled. The observed symptom was a segmentation fault on return from
4311 the function os_common_find_running_interfaces() in src/os.c.
4313 3. In the check_special_case() function in daemon.c I had used "errno" as an
4314 argument name, which causes warnings on some systems. This was basically a
4315 typo, since it was named "eno" in the comments!
4317 4. The code that waits for the clock to tick (at a resolution of some fraction
4318 of a second) so as to ensure message-id uniqueness was always waiting for
4319 at least one whole tick, when it could have waited for less. [This is
4320 almost certainly not relevant at current processor speeds, where it is
4321 unlikely to ever wait at all. But we try to future-proof.]
4323 5. The function that sleeps for a time interval that includes fractions of a
4324 second contained a race. It did not block SIGALRM between setting the
4325 timer, and suspending (a couple of lines later). If the interval was short
4326 and the sigsuspend() was delayed until after it had expired, the suspension
4327 never ended. On busy systems this could lead to processes getting stuck for
4330 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4331 process, before it forks any delivery processes. The open lookup caching
4332 mechanism meant that the open file or database connection was passed into
4333 the delivery process. The problem was that delivery processes always tidy
4334 up cached lookup data. This could cause a problem for the next delivery
4335 process started by the queue runner, because the external queue runner
4336 process does not know about the closure. So the next delivery process
4337 still has data in the lookup cache. In the case of a file lookup, there was
4338 no problem because closing a file descriptor in a subprocess doesn't affect
4339 the parent. However, if the lookup was caching a connection to a database,
4340 the connection was closed, and the second delivery process was likely to
4341 see errors such as "PGSQL: query failed: server closed the connection
4342 unexpectedly". The problem has been fixed by closing all cached lookups
4343 in a queue runner before running a delivery process.
4345 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4346 seem to have the "const" qualifier which it has on other OS. I've
4349 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4350 *of the same type* (client or server) with the same public name that an
4351 error should be diagnosed.
4353 9. When Exim looked up a host name for an IP address, but failed to find the
4354 original IP address when looking up the host name (a safety check), it
4355 output the message "<ip address> does not match any IP for NULL", which was
4356 confusing, to say the least. The bug was that the host name should have
4357 appeared instead of "NULL".
4359 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4360 user that is built into the binary, and the -C or -D options is used, root
4361 privilege is dropped before the configuration file is read. In addition,
4362 logging is switched to stderr instead of the normal log files. If the
4363 configuration then re-defines the Exim user, the unprivileged environment
4364 is probably not what is expected, so Exim logs a panic warning message (but
4367 However, if deliver_drop_privilege is set, the unprivileged state may well
4368 be exactly what is intended, so the warning has been cut out in that case,
4369 and Exim is allowed to try to write to its normal log files.
4375 1. smtp_return_error_details was not giving details for temporary sender
4376 or receiver verification errors.
4378 2. Diagnose a configuration error if two authenticators have the same public
4381 3. Exim used not to create the message log file for a message until the first
4382 delivery attempt. This could be confusing when incoming messages were held
4383 for policy or load reasons. The message log file is now created at the time
4384 the message is received, and an initial "Received" line is written to it.
4386 4. The automatically generated man page for command line options had a minor
4387 bug that caused no ill effects; however, a more serious problem was that
4388 the procedure for building the man page automatically didn't always
4389 operate. Consequently, release 4.20 contains an out-of-date version. This
4390 shouldn't happen again.
4392 5. When building Exim with embedded Perl support, the script that builds the
4393 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4394 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4396 6. The freeze_tell option was not being used for messages that were frozen on
4397 arrival, either by an ACL or by local_scan().
4399 7. Added the smtp_incomplete_transaction log selector.
4401 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4402 was accepting AUTH without a new EHLO.
4404 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4405 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4408 10. Logging of TCP/IP connections (when configured) now happens in the main
4409 daemon process instead of the child process, so that the TCP/IP connection
4410 count is more accurate (but it can never be perfect).
4412 11. The use of "drop" in a nested ACL was not being handled correctly in the
4413 outer ACL. Now, if condition failure induced by the nested "drop" causes
4414 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4415 or "require"), the connection is dropped.
4417 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4418 that yield "discard" can now be used with an "accept" or a "discard" verb,
4419 but an error is generated for any others (because I can't see a useful way
4420 to define what should happen).
4422 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4423 are now processed in the same way as lines in the Exim configuration file.
4424 In particular, continuation lines are supported.
4426 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4428 15. Added -ti meaning -t -i.
4430 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4431 domains, and warn by logging if others are found.
4433 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4434 returned from GIFCONF when Exim is trying to find the list of interfaces on
4435 a host. The code in os.c has been modified to copy each ifreq to an aligned
4436 structure in all cases.
4438 Also, in some cases, the returned ifreq's were being copied to a 'struct
4439 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4440 means the last couple of bytes of an IPv6 address could be chopped if the
4441 ifreq contained only a normal sockaddr (14 bytes storage).
4443 18. Named domain lists were not supported in the hosts_treat_as_local option.
4444 An entry such as +xxxx was not recognized, and was treated as a literal
4447 19. Ensure that header lines added by a DATA ACL are included in the reject log
4448 if the ACL subsequently rejects the message.
4450 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4451 MD5 (which is deprecated).
4453 21. When testing a filter file using -bf, Exim was writing a message when it
4454 took the sender from a "From " line in the message, but it was not doing so
4455 when it took $return_path from a Return-Path: header line. It now does.
4457 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4458 with a valid header line field name (a series of printing characters
4459 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4461 23. Changed "disc" in the source to "disk" to conform to the documentation and
4462 the book and for uniformity.
4464 24. Ignore Sendmail's -Ooption=value command line item.
4466 25. When execve() failed while trying to run a command in a pipe transport,
4467 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4468 could be confused with a return value of 69 from the command itself. This
4469 has been changed to 127, the value the shell returns if it is asked to run
4470 a non-existent command. The wording for the related log line suggests a
4471 non-existent command as the problem.
4473 26. If received_header_text expands to an empty string, do not add a Received:
4474 header line to the message. (Well, it adds a token one on the spool, but
4475 marks it "old" so that it doesn't get used or transmitted.)
4477 27. Installed eximstats 1.28 (addition of -nt option).
4479 28. There was no check for failure on the call to getsockname() in the daemon
4480 code. This can fail if there is a shortage of resources on the system, with
4481 ENOMEM, for example. A temporary error is now given on failure.
4483 29. Contrary to the C standard, it seems that in some environments, the
4484 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4485 program. Exim now does this explicitly; it affects the formatting of
4486 timestamps using strftime().
4488 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4489 complaints. I've now initialized all the variables, to avoid this.
4491 32. Header lines added by a system filter were not being "seen" during
4492 transport-time rewrites.
4494 33. The info_callback() function passed to OpenSSL is set up with type void
4495 (*)(SSL *, int, int), as described somewhere. However, when calling the
4496 function (actually a macro) that sets it up, the type void(*)() is
4497 expected. I've put in a cast to prevent warnings from picky compilers.
4499 34. If a DNS black list lookup found a CNAME record, but there were no A
4500 records associated with the domain it pointed at, Exim crashed.
4502 35. If a DNS black list lookup returned more than one A record, Exim ignored
4503 all but the first. It now scans all returned addresses if a particular IP
4504 value is being sought. In this situation, the contents of the
4505 $dnslist_value variable are a list of all the addresses, separated by a
4508 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4509 to accept a host name and all its aliases if the forward lookup for any of
4510 them yielded the IP address of the incoming connection. Now it accepts only
4511 those names whose forward lookup yields the correct IP address. Any other
4512 names are discarded. This closes a loophole whereby a rogue DNS
4513 administrator could create reverse DNS records to break through a
4514 wildcarded host restriction in an ACL.
4516 37. If a user filter or a system filter that ran in a subprocess used any of
4517 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4518 the wrong values were passed to the pipe command ($thisaddress had the
4519 value of $0, $0 had the value of $1, etc). This bug was introduced by
4520 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4522 38. Improved the line breaking for long SMTP error messages from ACLs.
4523 Previously, if there was no break point between 40 and 75 characters, Exim
4524 left the rest of the message alone. Two changes have been made: (a) I've
4525 reduced the minimum length to 35 characters; (b) if it can't find a break
4526 point between 35 and 75 characters, it looks ahead and uses the first one
4527 that it finds. This may give the occasional overlong line, but at least the
4528 remaining text gets split now.
4530 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4531 file descriptors might be low, and that setting 1000 would always raise it.
4532 It turns out that in some environments, the limit is already over 1000 and
4533 that lowering it causes trouble. So now Exim takes care not to decrease it.
4535 40. When delivering a message, the value of $return_path is set to $sender_
4536 address at the start of routing (routers may change the value). By an
4537 oversight, this default was not being set up when an address was tested by
4538 -bt or -bv, which affected the outcome if any router or filter referred to
4541 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4542 log only when "message" or "log_message" are set. However, if one of the
4543 conditions was an address verification, or a call to a nested ACL, the
4544 messages generated by the underlying test were being passed through. This
4545 no longer happens. The underlying message is available in $acl_verify_
4546 message for both "message" and "log_message" expansions, so it can be
4547 passed through if needed.
4549 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4550 new expansion $bh_ to give the encoded byte string without charset
4551 translation. Translation happens only if iconv() is available; HAVE_ICONV
4552 indicates this at build time. HEADERS_CHARSET gives the charset to
4553 translate to; headers_charset can change it in the configuration, and
4554 "headers charset" can change it in an individual filter file.
4556 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4557 that creates RFC 2047 encoded "words" labels them as that charset instead
4558 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4559 expansion operator; (ii) when Exim creates a From: line for a local
4560 message; (iii) when a header line is rewritten to include a "phrase" part.
4562 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4563 buggy, causing it to skip the first lines of messages whose message ID
4564 ended in 'D'. This would not have bitten before Exim release 4.14, because
4565 message IDs were unlikely to end in 'D' before then. The effect was to have
4566 incorrect size information for certain domains.
4568 45. #include "config.h" was missing at the start of the crypt16.c module. This
4569 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4572 46. If there was a timeout during a "random" callout check, Exim treated it as
4573 a failure of the random address, and carried on sending RSET and the real
4574 address. If the delay was just some slowness somewhere, the response to the
4575 original RCPT would be taken as a response to RSET and so on, causing
4576 mayhem of various kinds.
4578 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4579 when I implemented it. It didn't allow for the fact that some option values
4580 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4581 the right test for positive values.
4583 48. Domain names in DNS records are case-independent. Exim always looks them up
4584 in lower case. Some resolvers return domain names in exactly the case they
4585 appear in the zone file, that is, they may contain uppercase letters. Not
4586 all resolvers do this - some return always lower case. Exim was treating a
4587 change of case by a resolver as a change of domain, similar to a widening
4588 of a domain abbreviation. This triggered its re-routing code and so it was
4589 trying to route what was effectively the same domain again. This normally
4590 caused routing to fail (because the router wouldn't handle the domain
4591 twice). Now Exim checks for this case specially, and just changes the
4592 casing of the domain that it ultimately uses when it transmits the message
4595 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4598 50. If a filter generated a file delivery with a non-absolute name (possible if
4599 no home directory exists for the router), the forbid_file option was not
4602 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4603 the existing equality matching.
4605 52. Exim was using ints instead of ino_t variables in some places where it was
4606 dealing with inode numbers.
4608 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4609 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4610 called TMPDIR, and if it finds it is different, it changes its value.
4612 54. The smtp_printf() function is now made available to local_scan() so
4613 additional output lines can be written before returning. There is also an
4614 smtp_fflush() function to enable the detection of a dropped connection.
4615 The variables smtp_input and smtp_batched_input are exported to
4618 55. Changed the default runtime configuration: the message "Unknown user"
4619 has been removed from the ACL, and instead placed on the localuser router,
4620 using the cannot_route_message feature. This means that any verification
4621 failures that generate their own messages won't get overridden. Similarly,
4622 the "Unrouteable address" message that was in the ACL for unverifiable
4623 relay addresses has also been removed.
4625 56. Added hosts_avoid_esmtp to the smtp transport.
4627 57. The exicyclog script was not checking for the esoteric option
4628 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4629 will work only if exicyclog is run under the appropriate euid.
4631 58. Following a discussion on the list, the rules by which Exim recognises line
4632 endings on incoming messages have been changed. The -dropcr and drop_cr
4633 options are now no-ops, retained only for backwards compatibility. The
4634 following line terminators are recognized: LF CRLF CR. However, special
4635 processing applies to CR:
4637 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4638 nor a local message in the state where . is a terminator.
4640 (ii) If a bare CR is encountered in a header line, an extra space is added
4641 after the line terminator so as not to end the header. The reasoning
4642 behind this is that bare CRs in header lines are most likely either
4643 to be mistakes, or people trying to play silly games.
4645 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4646 was being incorrectly given as 18 bytes larger than it should have been.
4647 This is a VOB (very old bug).
4649 60. This may never have affected anything current, but just in case it has:
4650 When the local host is found other than at the start of a list of hosts,
4651 the local host, those with the same MX, and any that follow, are discarded.
4652 When the list in question was part of a longer list of hosts, the following
4653 hosts (not currently being processed) were also being discarded. This no
4654 longer happens. I'm not sure if this situation could ever has previously
4657 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4660 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4661 header. This is something that is recommended in a new Internet Draft, and
4662 is something that is documented as being done by Sendmail. There are two
4663 possible values. For messages generated by the autoreply transport, Exim
4666 Auto-Submitted: auto-replied
4668 whereas for all other generated messages (e.g. bounces) it adds
4670 Auto-Submitted: auto-generated
4672 63. The "personal" condition in filters now includes a test for the
4673 Auto-Submitted: header. If it contains the string "auto-" the message it
4674 not considered personal.
4676 64. Added rcpt_include_affixes as a generic transport option.
4678 65. Added queue_only_override (default true).
4680 66. Added the syslog_duplication option.
4682 67. If what should have been the first header line of a message consisted of
4683 a space followed by a colon, Exim was mis-interpreting it as a header line.
4684 It isn't of course - it is syntactically invalid and should therefore be
4685 treated as the start of the message body. The misbehaviour could have
4686 caused a number of strange effects, including loss of data in subsequent
4687 header lines, and spool format errors.
4689 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4690 client host had authenticated. This control can now be exercised by an ACL
4691 for more flexibility.
4693 69. By default, callouts do not happen when testing with -bh. There is now a
4694 variant, -bhc, which does actually run the callout code, including
4695 consulting and updating the callout cache.
4697 70. Added support for saslauthd authentication, courtesy of Alexander
4700 71. If statvfs() failed on the spool or log directories while checking their
4701 size for availability, Exim confusingly gave the error "space shortage".
4702 Furthermore, in debugging mode it crashed with a floating point exception.
4703 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4704 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4705 serious problem, Exim now writes to the main and panic logs when this
4706 happens, with details of the failure. It then refuses to accept the
4707 incoming message, giving the message "spool directory problem" or "log
4708 directory problem" with a 421 code for SMTP messages.
4710 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4711 0, 1, and 2 exist, because some OS complain for execs without them (see
4712 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4713 descriptors. However, the code omitted to check that the open succeeded,
4714 causing mysterious errors if for some reason the permissions on /dev/null
4715 got screwed. Now Exim writes a message to the main and panic logs, and
4716 bombs out if it can't open /dev/null.
4718 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4719 interact so that it is all more flexible. It is supposed to remain
4720 backwards compatible. Also added extra_local_interfaces.
4722 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4723 to bomb out with an assertion failure - to the client this appears as a
4724 connection drop. This problem occurs in the part of the code that was taken
4725 from the Samba project. Fortunately, the assertion is in a very simple
4726 function, so I have fixed this by reproducing the function inline in the
4727 one place where it is called, and arranging for authentication to fail
4728 instead of killing the process with assert().
4730 75. The SPA client code was not working when the server requested OEM rather
4731 than Unicode encoding.
4733 76. Added code to make require_files with a specific uid setting more usable in
4734 the case where statting the file as root fails - usually a non-root-mounted
4735 NFS file system. When this happens and the failure is EACCES, Exim now
4736 forks a subprocess and does the per-uid checking as the relevant uid.
4738 77. Added process_log_path.
4740 78. If log_file_path was not explicitly set, a setting of check_log_space or
4741 check_log_inodes was ignored.
4743 79. If a space check for the spool or log partitions fails, the incident is now
4744 logged. Of course, in the latter case the data may get lost...
4746 80. Added the %p formatting code to string_format() so that it can be used to
4747 print addresses in debug_print(). Adjusted all the address printing in the
4748 debugging in store.c to use %p rather than %d.
4750 81. There was a concern that a line of code in smtp_in.c could overflow a
4751 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4752 initially expressed, the concern was not well-founded, because trailing
4753 spaces are removed early. However, if the trailing spaces were followed by
4754 a NULL, they did not get removed, so the overflow was possible. Two fixes
4757 (a) I re-wrote the offending code in a cleaner fashion.
4758 (b) If an incoming SMTP command contains a NULL character, it is rejected
4761 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4762 runs initgroups(), so that if the Exim user is in any additional groups,
4763 they will be used during message reception.
4769 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob