1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.507 2007/05/11 08:50:42 tom 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
35 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
36 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
39 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
42 PH/02 In an ACL statement such as
44 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
46 if a client was not listed at all, or was listed with a value other than
47 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
48 the condition was not true (as it should be), so access was not denied.
49 The bug was that the ! inversion was incorrectly passed on to the second
50 item. This has been fixed.
52 PH/03 Added additional dnslists conditions == and =& which are different from
53 = and & when the dns lookup returns more than one IP address.
55 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
56 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
58 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
59 FSYNC, which compiles an option called disable_fsync that allows for
60 bypassing fsync(). The documentation is heavily laced with warnings.
62 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
64 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
65 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
66 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
67 including adding "make clean"; (3) Added -fPIC when compiling the test
68 dynamically loaded module, to get rid of a warning.
70 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
71 message fails, move_frozen_messages = true and ignore_bounce_errors_after
72 = 0s. The bug is otherwise harmless.
74 PH/07 There was a bug in the dovecot authenticator such that the value of
75 $auth1 could be overwritten, and so not correctly preserved, after a
76 successful authentication. This usually meant that the value preserved by
77 the server_setid option was incorrect.
79 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
81 PH/09 Installed PCRE release 7.0.
83 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
84 run for batched SMTP input. It is now run at the start of every message
85 in the batch. While fixing this I discovered that the process information
86 (output by running exiwhat) was not always getting set for -bs and -bS
87 input. This is fixed, and it now also says "batched" for BSMTP.
89 PH/11 Added control=no_pipelining.
91 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
92 patch, slightly modified), and move the expansion of helo_data till after
93 the connection is made in the smtp transport (so it can use these
96 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
98 PH/14 Added log_selector = +pid.
100 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
102 PH/16 Add ${if forany and ${if forall.
104 PH/17 Added dsn_from option to vary the From: line in DSNs.
106 PH/18 Flush SMTP output before performing a callout, unless control =
107 no_callout_flush is set.
109 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
110 was true (the default) a successful delivery failed to delete the retry
111 item, thus causing premature timeout of the address. The bug is now
114 PH/20 Added hosts_avoid_pipelining to the smtp transport.
116 PH/21 Long custom messages for fakedefer and fakereject are now split up
117 into multiline reponses in the same way that messages for "deny" and
118 other ACL rejections are.
120 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
121 with slight modification.
123 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
124 draft, changing the syntax and factoring some duplicate code".
126 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
127 for deliveries of the second and subsequent messages over the same SMTP
130 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
131 ${reduce, with only minor "tidies".
133 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
135 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
136 expansion side effects.
138 PH/27 When a message times out after an over-quota error from an Exim-imposed
139 quota, the bounce message says "mailbox is full". This message was not
140 being given when it was a system quota that was exceeded. It now should
143 MH/03 Made $recipients available in local_scan(). local_scan() already has
144 better access to the recipient list through recipients_list[], but
145 $recipients can be useful in postmaster-provided expansion strings.
147 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
148 in the case of a MAIL command with additional options following the
149 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
150 were accidentally chopped off.
152 PH/29 SMTP synchronization checks are implemented when a command is read -
153 there is a check that no more input is waiting when there shouldn't be
154 any. However, for some commands, a delay in an ACL can mean that it is
155 some time before the response is written. In this time, more input might
156 arrive, invalidly. So now there are extra checks after an ACL has run for
157 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
158 pipelining has not been advertised.
160 PH/30 MH's patch to allow iscntrl() characters to be list separators.
162 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
163 returned in the SMTP response when smtp_return_error_details was false.
166 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
167 instead of the C I/O that was originally supplied, because problems were
170 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
171 Exim which did not show up earlier: it was assuming that a call to
172 SSL_CTX_set_info_callback() might give an error value. In fact, there is
173 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
174 was a macro that became an assignment, so it seemed to work. This has
175 changed to a proper function call with a void return, hence the compile
176 error. Exim's code has been fixed.
178 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
181 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
183 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
185 PH/37 If a message is not accepted after it has had an id assigned (e.g.
186 because it turns out to be too big or there is a timeout) there is no
187 "Completed" line in the log. When some messages of this type were
188 selected by exigrep, they were listed as "not completed". Others were
189 picked up by some special patterns. I have improved the selection
190 criteria to be more general.
192 PH/38 The host_find_failed option in the manualroute router can now be set
193 to "ignore", to completely ignore a host whose IP address cannot be
194 found. If all hosts are ignored, the behaviour is controlled by the new
195 host_all_ignored option.
197 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
198 homing or because of multiple MX records with /mx) generated more than
199 one IP address, and the following item turned out to be the local host,
200 all the secondary addresses of the first item were incorrectly removed
201 from the list, along with the local host and any following hosts (which
202 is what is supposed to happen).
204 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
205 whoever called Exim into the -H file. In the case of the daemon it was
206 behaving confusingly. When first started, it used values for whoever
207 started the daemon, but after a SIGHUP it used the Exim user (because it
208 calls itself on a restart). I have changed the code so that it now always
211 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
212 message are rejected with the same error (e.g. no authentication or bad
213 sender address), and a DATA command is nevertheless sent (as can happen
214 with PIPELINING or a stupid MUA), the error message that was given to the
215 RCPT commands is included in the rejection of the DATA command. This is
216 intended to be helpful for MUAs that show only the final error to their
219 PH/42 Another patch from the Sieve maintainer.
221 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
222 Eximstats - Fixed some broken HTML links and added missing column headers
224 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
225 columns for Rejects, Temp Rejects, Ham, and Spam rows.
227 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
229 PH/43 Yet another patch from the Sieve maintainer.
231 PH/44 I found a way to check for a TCP/IP connection going away before sending
232 the response to the final '.' that terminates a message, but only in the
233 case where the client has not sent further data following the '.'
234 (unfortunately, this is allowed). However, in many cases there won't be
235 any further data because there won't be any more messages to send. A call
236 to select() can be used: if it shows that the input is "ready", there is
237 either input waiting, or the socket has been closed. An attempt to read
238 the next input character can distinguish the two cases. Previously, Exim
239 would have sent an OK response which the client would never have see.
240 This could lead to message repetition. This fix should cure that, at
241 least in a lot of common cases.
243 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
244 advertised in response to EHLO.
250 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
251 fixed by 4.65/MH/01 (is this a record?) are fixed:
253 (i) An empty string was always treated as zero by the numeric comparison
254 operators. This behaviour has been restored.
256 (ii) It is documented that the numeric comparison operators always treat
257 their arguments as decimal numbers. This was broken in that numbers
258 starting with 0 were being interpreted as octal.
260 While fixing these problems I realized that there was another issue that
261 hadn't been noticed. Values of message_size_limit (both the global option
262 and the transport option) were treated as octal if they started with 0.
263 The documentation was vague. These values are now always treated as
264 decimal, and I will make that clear in the documentation.
270 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
271 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
274 MH/01 Don't check that the operands of numeric comparison operators are
275 integers when their expansion is in "skipping" mode (fixes bug
276 introduced by 4.64-PH/07).
278 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
279 child addresses, Exim now panics and dies. Previously, because the count
280 is held in a short int, deliveries were likely to be lost. As such a
281 large number of recipients for a single message is ridiculous
282 (performance will be very, very poor), I have chosen to impose a limit
283 rather than extend the field.
289 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
290 leftover -K file (the existence of which was triggered by #402).
291 While we were at it, introduced process PID as part of the -K
292 filename. This should rule out race conditions when creating
295 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
296 processing considerably. Previous code took too long for large mails,
297 triggering a timeout which in turn triggers #401.
299 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
300 in the DK code in transports.c. sendfile() is not really portable,
301 hence the _LINUX specificness.
303 TF/01 In the add_headers option to the mail command in an Exim filter,
304 there was a bug that Exim would claim a syntax error in any
305 header after the first one which had an odd number of characters
308 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
309 callout verification, Exim cached a "reject" for the entire domain. This
310 is correct for most verifications, but it is not correct for a recipient
311 verification with use_sender or use_postmaster set, because in that case
312 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
313 case of MAIL FROM:<> rejection from other early rejections (e.g.
314 rejection of HELO). When verifying a recipient using a non-null MAIL
315 address, the cache is ignored if it shows MAIL FROM:<> rejection.
316 Whatever the result of the callout, the value of the domain cache is
317 left unchanged (for any other kind of callout, getting as far as trying
318 RCPT means that the domain itself is ok).
320 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
323 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
324 manifest itself as EPIPE rather than ECONNECT. When tidying away a
325 session, the daemon ignores ECONNECT errors and logs others; it now
326 ignores EPIPE as well.
328 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
329 (quoted-printable decoding).
331 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
332 later the small subsequent patch to fix an introduced bug.
334 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
336 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
338 PH/08 An error is now given if message_size_limit is specified negative.
340 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
341 to be given (somewhat) arbitrary names.
343 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
346 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
347 miscellaneous code fixes
349 PH/10 Added the log_reject_target ACL modifier to specify where to log
352 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
353 hostname. This is wrong, because it relates to the incoming message (and
354 probably the interface on which it is arriving) and not to the outgoing
355 callout (which could be using a different interface). This has been
356 changed to use the value of the helo_data option from the smtp transport
357 instead - this is what is used when a message is actually being sent. If
358 there is no remote transport (possible with a router that sets up host
359 addresses), $smtp_active_hostname is used.
361 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
362 tweaks were necessary in order to get it to work (see also 21 below):
363 (a) The code assumed that strncpy() returns a negative number on buffer
364 overflow, which isn't the case. Replaced with Exim's string_format()
366 (b) There were several signed/unsigned issues. I just did the minimum
367 hacking in of casts. There is scope for a larger refactoring.
368 (c) The code used strcasecmp() which is not a standard C function.
369 Replaced with Exim's strcmpic() function.
370 (d) The code set only $1; it now sets $auth1 as well.
371 (e) A simple test gave the error "authentication client didn't specify
372 service in request". It would seem that Dovecot has changed its
373 interface. Fortunately there's a specification; I followed it and
374 changed what the client sends and it appears to be working now.
376 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
379 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
380 address A is aliased to B and C, where B exists and C does not. Without
381 -v the output is "A verified" because verification stops after a
382 successful redirection if more than one address is generated. However,
383 with -v the child addresses are also verified. Exim was outputting "A
384 failed to verify" and then showing the successful verification for C,
385 with its parentage. It now outputs "B failed to verify", showing B's
386 parentage before showing the successful verification of C.
388 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
389 look up a TXT record in a specific list after matching in a combined
392 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
393 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
394 they consult the DNS. I had assumed they would set it the way they
395 wanted; and indeed my experiments on Linux seem to show that in some
396 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
397 To be on the safe side, however, I have now made the interface to
398 host_find_byname() similar to host_find_bydns(), with an argument
399 containing the DNS resolver options. The host_find_byname() function now
400 sets these options at its start, just as host_find_bydns() does. The smtp
401 transport options dns_qualify_single and dns_search_parents are passed to
402 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
403 of host_find_byname() use the default settings of RES_DEFNAMES
404 (qualify_single) but not RES_DNSRCH (search_parents).
406 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
407 spool_read_header() do less string testing, by means of a preliminary
408 switch on the second character of optional "-foo" lines. (This is
409 overdue, caused by the large number of possibilities that now exist.
410 Originally there were few.) While I was there, I also converted the
411 str(n)cmp tests so they don't re-test the leading "-" and the first
412 character, in the hope this might squeeze out yet more improvement.
414 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
415 flag allowing group syntax was set by the header_syntax check but not
416 turned off, possible causing trouble later; (2) The flag was not being
417 set at all for the header_verify test, causing "group"-style headers to
418 be rejected. I have now set it in this case, and also caused header_
419 verify to ignore an empty address taken from a group. While doing this, I
420 came across some other cases where the code for allowing group syntax
421 while scanning a header line wasn't quite right (mostly, not resetting
422 the flag correctly in the right place). These bugs could have caused
423 trouble for malformed header lines. I hope it is now all correct.
425 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
426 with the "reply" argument non-NULL. The code, however (which originally
427 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
428 but it didn't always do it. This confused somebody who was copying the
429 code for some other use. I have removed all the tests.
431 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
432 feature that was used to support insecure browsers during the U.S. crypto
433 embargo. It requires special client support, and Exim is probably the
434 only MTA that supported it -- and would never use it because real RSA is
435 always available. This code has been removed, because it had the bad
436 effect of slowing Exim down by computing (never used) parameters for the
437 RSA_EXPORT functionality.
439 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
440 authenticator to fail if there's a tab character in the incoming data
441 (there should never be unless someone is messing about, as it's supposed
442 to be base64-encoded). Also added, on Timo's advice, the "secured" option
443 if the connection is using TLS or if the remote IP is the same as the
444 local IP, and the "valid-client-cert option" if a client certificate has
447 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
448 authenticators. This can be used for authorization after authentication
449 succeeds. (In the case of plaintext, it servers for both authentication
452 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
453 if any retry times were supplied.
455 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
456 connection, where there is no client IP address to check. In this
457 situation, the verify now always succeeds.
459 PH/25 Applied John Jetmore's -Mset patch.
461 PH/26 Added -bem to be like -Mset, but loading a message from a file.
463 PH/27 In a string expansion for a processed (not raw) header when multiple
464 headers of the same name were present, leading whitespace was being
465 removed from all of them, but trailing whitespace was being removed only
466 from the last one. Now trailing whitespace is removed from each header
467 before concatenation. Completely empty headers in a concatenation (as
470 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
471 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
473 PH/29 [Removed. This was a change that I later backed out, and forgot to
474 correct the ChangeLog entry (that I had efficiently created) before
475 committing the later change.]
477 PH/30 Exim was sometimes attempting to deliver messages that had suffered
478 address errors (4xx response to RCPT) over the same connection as other
479 messages routed to the same hosts. Such deliveries are always "forced",
480 so retry times are not inspected. This resulted in far too many retries
481 for the affected addresses. The effect occurred only when there were more
482 hosts than the hosts_max_try setting in the smtp transport when it had
483 the 4xx errors. Those hosts that it had tried were not added to the list
484 of hosts for which the message was waiting, so if all were tried, there
485 was no problem. Two fixes have been applied:
487 (i) If there are any address or message errors in an SMTP delivery, none
488 of the hosts (tried or untried) are now added to the list of hosts
489 for which the message is waiting, so the message should not be a
490 candidate for sending over the same connection that was used for a
491 successful delivery of some other message. This seems entirely
492 reasonable: after all the message is NOT "waiting for some host".
493 This is so "obvious" that I'm not sure why it wasn't done
494 previously. Hope I haven't missed anything, but it can't do any
495 harm, as the worst effect is to miss an optimization.
497 (ii) If, despite (i), such a delivery is accidentally attempted, the
498 routing retry time is respected, so at least it doesn't keep
499 hammering the server.
501 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
502 in ${readsocket because some servers need this prod.
504 PH/32 Added some extra debug output when updating a wait-xxx database.
506 PH/33 The hint "could be header name not terminated by colon", which has been
507 given for certain expansion errors for a long time, was not being given
508 for the ${if def:h_colon_omitted{... case.
510 PH/34 The spec says: "With one important exception, whenever a domain list is
511 being scanned, $domain contains the subject domain." There was at least
512 one case where this was not true.
514 PH/35 The error "getsockname() failed: connection reset by peer" was being
515 written to the panic log as well as the main log, but it isn't really
516 panic-worthy as it just means the connection died rather early on. I have
517 removed the panic log writing for the ECONNRESET error when getsockname()
520 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
521 runs only) independently of the message's sender address. This meant
522 that, if the 4xx error was in fact related to the sender, a different
523 message to the same recipient with a different sender could confuse
524 things. In particualar, this can happen when sending to a greylisting
525 server, but other circumstances could also provoke similar problems.
526 I have changed the default so that the retry time for these errors is now
527 based a combination of the sender and recipient addresses. This change
528 can be overridden by setting address_retry_include_sender=false in the
531 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
532 remote server are returned as part of bounce messages. This was not
533 happening for LMTP over a pipe (the lmtp transport), but now it is the
534 same for both kinds of LMTP.
536 PH/38 Despite being documented as not happening, Exim was rewriting addresses
537 in header lines that were in fact CNAMEs. This is no longer the case.
539 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
540 and queue runs started by the daemon processed all messages. This has
541 been fixed so that -R and -S can now usefully be given with -q<time>.
543 PH/40 Import PCRE release 6.7 (fixes some bugs).
545 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
547 PH/42 Give an error if -q is specified more than once.
549 PH/43 Renamed the variables $interface_address and $interface_port as
550 $received_ip_address and $received_port, to make it clear that these
551 values apply to message reception, and not to the outgoing interface when
552 a message is delivered. (The old names remain recognized, of course.)
554 PH/44 There was no timeout on the connect() call when using a Unix domain
555 socket in the ${readsocket expansion. There now is.
557 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
558 be meaningful with "accept".
560 SC/01 Eximstats V1.43
561 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
563 SC/02 Eximstats V1.44
564 Use a glob alias rather than an array ref in the generated
565 parser. This improves both readability and performance.
567 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
568 Collect SpamAssassin and rejection statistics.
569 Don't display local sender or destination tables unless
570 there is data to show.
571 Added average volumes into the top table text output.
573 SC/04 Eximstats V1.46
574 Collect data on the number of addresses (recipients)
575 as well as the number of messages.
577 SC/05 Eximstats V1.47
578 Added 'Message too big' to the list of mail rejection
579 reasons (thanks to Marco Gaiarin).
581 SC/06 Eximstats V1.48
582 Mainlog lines which have GMT offsets and are too short to
583 have a flag are now skipped.
585 SC/07 Eximstats V1.49 (Alain Williams)
586 Added the -emptyok flag.
588 SC/08 Eximstats V1.50
589 Fixes for obtaining the IP address from reject messages.
591 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
592 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
593 whitesspace changes from 4.64-PH/27
595 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
598 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
599 are found, allow negative numbers in numeric criteria)
601 JJ/06 exipick.20061117.2, added new $message_body_missing variable
603 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
604 to match changes made in 4.64-PH/43
606 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
608 PH/47 Put in an explicit test for a DNS lookup of an address record where the
609 "domain" is actually an IP address, and force a failure. This locks out
610 those revolvers/nameservers that support "A-for-A" lookups, in
611 contravention of the specifications.
613 PH/48 When a host name was looked up from an IP address, and the subsequent
614 forward lookup of the name timed out, the host name was left in
615 $sender_host_name, contrary to the specification.
617 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
618 restricted to single-key lookups, Exim was not diagnosing an error if
619 * or *@ was used with a query-style lookup.
621 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
623 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
624 long ago, but noone interested enough thought of it. Let's just say that
625 the "1.1" means that there are some new functions that weren't there at
626 some point in the past.
628 PH/51 Error processing for expansion failure of helo_data from an smtp
629 transport during callout processing was broken.
631 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
632 tested/used via the -bh/-bhc/-bs options.
634 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
635 bug, fixed in subsequent PCRE releases).
637 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
638 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
640 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
646 SC/01 Use a glob alias rather than an array ref in eximstats generated
647 parser. This improves both readability and performance.
649 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
650 Don't display local sender or destination tables in eximstats unless
651 there is data to show.
652 Added average volumes into the eximstats top table text output.
654 SC/03 Collect data on the number of addresses (recipients) as well
655 as the number of messages in eximstats.
657 TF/01 Correct an error in the documentation for the redirect router. Exim
658 does (usually) call initgroups() when daemonizing.
660 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
661 with consistent privilege compared to when running as a daemon.
663 TF/03 Note in the spec that $authenticated_id is not set for local
664 submissions from trusted users.
666 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
667 Thanks to Dean Brooks <dean@iglou.com> for the patch.
669 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
670 by adding some example configuration directives to the default
671 configuration file. A little bit of work is required to uncomment the
672 directives and define how usernames and passwords are checked, but
673 there is now a framework to start from.
675 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
676 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
677 without this. I don't know how relevant this is to other LDAP libraries.
679 PH/02 Add the verb name to the "unknown ACL verb" error.
681 PH/03 Magnus Holmgren's patch for filter_prepend_home.
683 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
685 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
686 directory not expanded when it should be if an expanded home directory
687 was set for the address (which is overridden by the transport).
689 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
692 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
693 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
694 because it is too late at that time, and has no effect.
696 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
697 security issue with \' (bugzilla #107). I could not use the
698 PQescapeStringConn() function, because it needs a PGconn value as one of
701 PH/08 When testing addresses using -bt, indicate those final addresses that
702 are duplicates that would not cause an additional delivery. At least one
703 person was confused, thinking that -bt output corresponded to deliveries.
704 (Suppressing duplicates isn't a good idea as you lose the information
705 about possibly different redirections that led to the duplicates.)
707 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
708 systems where poll() doesn't work, in particular OS X.
710 PH/10 Added more information to debugging output for retry time not reached.
712 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
713 operations in malware.c.
715 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
718 PH/13 If write_rejectlog was set false when logging was sent to syslog with
719 syslog_duplication set false, log lines that would normally be written
720 both the the main log and to the reject log were not written to syslog at
723 PH/14 In the default configuration, change the use of "message" in ACL warn
724 statements to "add_header".
726 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
727 not followed by a command (e.g. "seen endif").
729 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
730 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
733 PH/17 Added extra conditions to the default value of delay_warning_condition
737 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
738 { match{$h_precedence:}{(?i)bulk|list|junk} } \
739 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
742 The Auto-Submitted: and various List- headers are standardised, whereas I
743 don't think Precedence: ever was.
745 PH/18 Refactored debugging code in route_finduser() to show more information,
746 in particular, the error code if getpwnam() issues one.
748 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
749 This is apparently needed in addition to the PH/07 change above to avoid
750 any possible encoding problems.
752 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
753 but not after initializing Perl.
755 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
756 output them only if debugging. By default they are written stderr,
757 apparently, which is not desirable.
759 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
762 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
765 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
767 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
768 authenticated or an ident call has been made. Suppress the default
769 values for $authenticated_id and $authenticated_sender (but permit -oMai
770 and -oMas) when testing with -bh.
772 PH/24 Re-jigged the order of the tests in the default configuration so that the
773 tests for valid domains and recipients precede the DNS black list and CSA
774 tests, on the grounds that those ones are more expensive.
776 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
777 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
778 This bug exists in every version of Exim that I still have, right back to
781 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
782 However, an attempt to turn on case-sensitivity in a regex key by
783 including (?-i) didn't work because the subject string was already
784 lowercased, and the effects were non-intuitive. It turns out that a
785 one-line patch can be used to allow (?-i) to work as expected.
791 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
792 other effects) broke the use of negated acl sub-conditions.
794 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
797 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
798 "Deny" causes Exim to reject the incoming connection with a 554 error.
799 Unfortunately, if there is a major crisis, such as a disk failure,
800 tcp-wrappers gives "deny", whereas what one would like would be some
801 kind of temporary error. A kludge has been added to help with this.
802 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
803 554 error is used if errno is still zero or contains ENOENT (which occurs
804 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
807 PH/03 Add -lutil to the default FreeBSD LIBS setting.
809 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
810 errors. Otherwise a message that provokes a temporary error (when other
811 messages do not) can cause a whole host to time out.
813 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
814 addresses were routed directly to files or pipes from a redirect router.
815 File deliveries just didn't batch; pipe deliveries might have suffered
818 PH/06 A failure to get a lock for a hints database would erroneously always say
819 "Failed to get write lock", even when it was really a read lock.
821 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
822 of 0600. This has been changed to use the value of the lockfile_mode
823 option (which defaults to 0600).
825 PH/08 Applied small patch from the Sieve maintainer.
827 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
828 folder from quota calculations, a direct delivery into this folder messed
829 up the contents of the maildirsize file. This was because the regex was
830 used only to exclude .Trash (or whatever) when the size of the mailbox
831 was calculated. There was no check that a delivery was happening into an
832 excluded directory. This bug has been fixed by ignoring all quota
833 processing for deliveries into excluded directories.
835 PH/10 Added the maildirfolder_create_regex option to appendfile.
841 PH/01 The code for finding all the local interface addresses on a FreeBSD
842 system running IPv6 was broken. This may well have applied to all BSD
843 systems, as well as to others that have similar system calls. The broken
844 code found IPv4 interfaces correctly, but gave incorrect values for the
845 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
846 that it would not match correctly against @[] and not recognize the IPv6
849 PH/02 The ipliteral router was not recognizing addresses of the form user@
850 [ipv6:....] because it didn't know about the "ipv6:" prefix.
852 PH/03 Added disable_ipv6.
854 PH/04 Changed $reply_address to use the raw form of the headers instead of the
855 decoded form, because it is most often used to construct To: headers
856 lines in autoreplies, and the decoded form may well be syntactically
857 invalid. However, $reply_address has leading white space removed, and all
858 newlines turned into spaces so that the autoreply transport does not
861 PH/05 If group was specified without a user on a router, and no group or user
862 was specified on a transport, the group from the router was ignored.
864 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
865 for visible compile-time settings that can be used to change these
866 numbers, for those that want even more. Backwards compatibility with old
867 spool files has been maintained. However, going back to a previous Exim
868 release will lost any variables that are in spool files.
870 PH/07 Two small changes when running in the test harness: increase delay when
871 passing a TCP/IP connection to a new process, in case the original
872 process has to generate a bounce, and remove special handling of
873 127.0.0.2 (sic), which is no longer necessary.
875 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
876 be the same on different OS.
878 PH/09 Moved a debug statement in filter processing to avoid a race problem when
881 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
882 whether --show-vars was specified or not
884 JJ/02 exipick: Added support for new ACL variable spool format introduced
887 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
888 syntactically invalid From: or Reply-to: line, and a filter used this to
889 generate an autoreply, and therefore failed to obtain an address for the
890 autoreply, Exim could try to deliver to a non-existent relative file
891 name, causing unrelated and misleading errors. What now happens is that
892 it logs this as a hard delivery error, but does not attempt to create a
895 PH/11 The exinext utility has a -C option for testing purposes, but although
896 the given file was scanned by exinext itself; it wasn't being passed on
897 when Exim was called.
899 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
900 an end-of-file indication when reading a command response.
902 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
903 compiled. In many other places in Exim, IPv6 addresses are always
904 recognized, so I have changed this. It also means that IPv4 domain
905 literals of the form [IPV4:n.n.n.n] are now always recognized.
907 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
908 used if the router is not running as root, for example, when verifying at
909 ACL time, or when using -bh. The debugging output from this situation was
910 non-existent - all you got was a failure to exec. I have made two
913 (a) Failures to set uid/gid, the current directory, or a process leader
914 in a subprocess such as that created by queryprogram now generate
915 suitable debugging ouput when -d is set.
917 (b) The queryprogram router detects when it is not running as root,
918 outputs suitable debugging information if -d is set, and then runs
919 the subprocess without attempting to change uid/gid.
921 PH/15 Minor change to Makefile for building test_host (undocumented testing
924 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
925 additional section of a DNS packet that returns MX or SRV records.
926 Instead, it always explicitly searches for A/AAAA records. This avoids
927 major problems that occur when a DNS server includes only records of one
928 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
929 fixed another bug: if SRV records were looked up and the corresponding
930 address records were *not* found in the additional section, the port
931 values from the SRV records were lost.
933 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
934 using the correct key (the original address) when searching the retry
935 rules in order to find which one to use for generating the retry hint.
937 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
938 adding the default one. Similarly, if it contains a Reply-To: header, the
939 errors_reply_to option, if set, is not used.
941 PH/19 When calculating a retry time, Exim used to measure the "time since
942 failure" by looking at the "first failed" field in the retry record. Now
943 it does not use this if it is later than than the arrival time of the
944 message. Instead it uses the arrival time. This makes for better
945 behaviour in cases where some deliveries succeed, thus re-setting the
946 "first failed" field. An example is a quota failure for a huge message
947 when small messages continue to be delivered. Without this change, the
948 "time since failure" will always be short, possible causing more frequent
949 delivery attempts for the huge message than are intended.
950 [Note: This change was subsequently modified - see PH/04 for 4.62.]
952 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
953 $1, $2, $3) because the numerical variables can be reset during some
954 expansion items (e.g. "match"), thereby losing the authentication data.
956 PH/21 Make -bV show the size of off_t variables so that the test suite can
957 decide whether to run tests for quotas > 2G.
959 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
960 mailbox_size, and mailbox_filecount in the appendfile transport. If a
961 filecount value is greater than 2G or if a quota value is greater than 2G
962 on a system where the size of off_t is not greater than 4, a panic error
965 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
966 never match. The debug and -bh output now contains an explicit error
967 message indicating a malformed IPv4 address or mask.
969 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
970 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
973 PH/25 Do not write to syslog when running in the test harness. The only
974 occasion when this arises is a failure to open the main or panic logs
975 (for which there is an explicit test).
977 PH/26 Added the /no_tell option to "control=freeze".
979 PH/27 If a host name lookup failed very early in a connection, for example, if
980 the IP address matched host_lookup and the reverse lookup yielded a name
981 that did not have a forward lookup, an error message of the form "no IP
982 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
983 could be logged. Now it outputs the IP address instead of "NULL".
985 PH/28 An enabling patch from MH: add new function child_open_exim2() which
986 allows the sender and the authenticated sender to be set when
987 submitting a message from within Exim. Since child_open_exim() is
988 documented for local_scan(), the new function should be too.
990 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
991 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
992 results in an empty string is now treated as unset.
994 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
996 PH/31 Added sender_verify_fail logging option.
998 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
999 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1000 tidied the source and removed it altogether.
1002 PH/33 When a queue run was abandoned because the load average was too high, a
1003 log line was always written; now it is written only if the queue_run log
1004 selector is set. In addition, the log line for abandonment now contains
1005 information about the queue run such as the pid. This is always present
1006 in "start" and "stop" lines but was omitted from the "abandon" line.
1008 PH/34 Omit spaces between a header name and the colon in the error message that
1009 is given when verify = headers_syntax fails (if there are lots of them,
1010 the message gets confusing).
1012 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1013 names, as there are now some PTR records that contain slashes. This check
1014 is only to protect against broken name servers that fall over on strange
1015 characters, so the fact that it applies to all lookups doesn't matter.
1017 PH/36 Now that the new test suite is complete, we can remove some of the
1018 special code in Exim that was needed for the old test suite. For example,
1019 sorting DNS records because real resolvers return them in an arbitrary
1020 order. The new test suite's fake resolver always returns records in the
1023 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1024 bounces) except when queue_only is set, to avoid logging races between
1025 the different processes.
1027 PH/38 Panic-die if .include specifies a non-absolute path.
1029 PH/39 A tweak to the "H" retry rule from its user.
1031 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1032 a label. They prevented compilation on older perls.
1034 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1035 a warning to be raised on newish perls.
1037 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1038 on queue. Changes to match documented behaviour of showing count of
1039 messages matching specified criteria.
1041 PH/40 Changed the default ident timeout from 30s to 5s.
1043 PH/41 Added support for the use of login_cap features, on those BSD systems
1044 that have them, for controlling the resources used by pipe deliveries.
1046 PH/42 The content-scanning code uses fopen() to create files in which to put
1047 message data. Previously it was not paying any attention to the mode of
1048 the files. Exim runs with umask(0) because the rest of the code creates
1049 files with open(), and sets the required mode explicitly. Thus, these
1050 files were ending up world-writeable. This was not a big issue, because,
1051 being within the spool directory, they were not world-accessible. I have
1052 created a function called modefopen, which takes an additional mode
1053 argument. It sets umask(777), creates the file, chmods it to the required
1054 mode, then resets the umask. All the relevant calls to fopen() in the
1055 content scanning code have been changed to use this function.
1057 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1058 to 24 hours. This avoids potential overflow problems when processing G
1059 and H retry rules. I suspect nobody ever tinkers with this value.
1061 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1063 PH/45 When the plaintext authenticator is running as a client, the server's
1064 challenges are checked to ensure they are valid base64 strings. By
1065 default, the authentication attempt is cancelled if an invalid string is
1066 received. Setting client_ignore_invalid_base64 true ignores these errors.
1067 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1068 they are received. Thus, the responses can be made to depend on the
1069 challenges. If an invalid string is ignored, an empty string is placed in
1072 PH/46 Messages that are created by the autoreply transport now contains a
1073 References: header, in accordance with RFCs 2822 and 3834.
1075 PH/47 Added authenticated_sender_force to the smtp transport.
1077 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1079 PH/49 Installed latest patch from the Sieve maintainer.
1081 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1082 was also set, the appendfile transport was unnecessarily scanning a
1083 directory of message files (e.g. for maildir delivery) to find the count
1084 of files (along with the size), even though it did not need this
1085 information. It now does the scan only if it needs to find either the
1086 size of the count of files.
1088 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1090 PH/52 Two bugs concerned with error handling when the smtp transport is
1093 (i) Exim was not creating retry information for temporary errors given
1094 for individual recipients after the DATA command when the smtp transport
1095 was used in LMTP mode. This meant that they could be retried too
1096 frequently, and not timed out correctly.
1098 (ii) Exim was setting the flag that allows error details to be returned
1099 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1100 recipients that were returned after the DATA command.
1102 PH/53 This is related to PH/52, but is more general: for any failing address,
1103 when detailed error information was permitted to be returned to the
1104 sender, but the error was temporary, then after the final timeout, only
1105 "retry timeout exceeded" was returned. Now it returns the full error as
1106 well as "retry timeout exceeded".
1108 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1109 do this, and (what is worse) MTAs that accept it.
1111 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1112 will now be deprecated.
1114 PH/56 New os.c-cygwin from the Cygwin maintainer.
1116 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1117 formats (previously only available in exim formats via -bpr, -bpru,
1118 and -bpra. Now also available in native and exiqgrep formats)
1120 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1121 with very large, slow to parse queues
1123 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1125 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1127 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1128 responses to authentication challenges, though it was showing the
1129 challenges; (ii) I've removed the CR characters from the debug output for
1132 PH/58 Allow for the insertion of a newline as well as a space when a string
1133 is turned into more than one encoded-word during RFC 2047 encoding. The
1134 Sieve code now uses this.
1136 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1137 data_4xx, lost_connection, tls_required.
1139 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1140 message was being sent as an SMTP response.
1142 PH/61 Add -l and -k options to exicyclog.
1144 PH/62 When verifying, if an address was redirected to one new address, so that
1145 verification continued, and the new address failed or deferred after
1146 having set something in $address_data, the value of $address_data was not
1147 passed back to the ACL. This was different to the case when no
1148 redirection occurred. The value is now passed back in both cases.
1150 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1151 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1152 use login_cap.h, so on its own it isn't the distinguishing feature. The
1153 new name refers directly to the setclassresources() function.
1155 PH/65 Added configuration files for NetBSD3.
1157 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1159 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1160 is preferred over IPv4.
1162 PH/68 The bounce_return_message and bounce_return_body options were not being
1163 honoured for bounces generated during the reception of non-SMTP messages.
1164 In particular, this applied to messages rejected by the ACL. This bug has
1165 been fixed. However, if bounce_return_message is true and bounce_return_
1166 body is false, the headers that are returned for a non-SMTP message
1167 include only those that have been read before the error was detected.
1168 (In the case of an ACL rejection, they have all been read.)
1170 PH/69 The HTML version of the specification is now built in a directory called
1171 spec_html instead of spec.html, because the latter looks like a path with
1172 a MIME-type, and this confuses some software.
1174 PH/70 Catch two compiler warnings in sieve.c.
1176 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1177 function verify_get_ident() calls ip_connect() to connect a socket, but
1178 if the "connect()" function timed out, ip_connect() used to close the
1179 socket. However, verify_get_ident() also closes the socket later, and in
1180 between Exim writes to the log, which may get opened at this point. When
1181 the socket was closed in ip_connect(), the log could get the same file
1182 descriptor number as the socket. This naturally causes chaos. The fix is
1183 not to close the socket in ip_connect(); the socket should be closed by
1184 the function that creates it. There was only one place in the code where
1185 this was missing, in the iplookup router, which I don't think anybody now
1186 uses, but I've fixed it anyway.
1188 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1189 well as to direct DNS lookups. Otherwise the handling of names in host
1190 lists is inconsistent and therefore confusing.
1196 PH/01 Two changes to the default runtime configuration:
1198 (1) Move the checks for relay_from_hosts and authenticated clients from
1199 after to before the (commented out) DNS black list checks.
1201 (2) Add control=submission to the relay_from_hosts and authenticated
1202 clients checks, on the grounds that messages accepted by these
1203 statements are most likely to be submissions.
1205 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1207 (1) Generate an error if the third argument for the ${prvs expansion is
1210 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1213 (3) Reset the variables that are obtained from the first argument of
1214 ${prvscheck and used in the second argument before leaving the code,
1215 because their memory is reclaimed, so using them afterwards may do
1218 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1219 one (it's much easier than Tom thought :-).
1221 (5) Because of (4), we can now allow for the use of $prvscheck_result
1222 inside the third argument.
1224 PH/03 For some reason, the default setting of PATH when running a command from
1225 a pipe transport was just "/usr/bin". I have changed it to
1228 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1229 anything to be listed in the output from -bV.
1231 PH/05 When a filter generated an autoreply, the entire To: header line was
1232 quoted in the delivery log line, like this:
1234 => >A.N.Other <ano@some.domain> <original@ddress> ...
1236 This has been changed so that it extracts the operative address. There
1237 may be more than one such address. If so, they are comma-separated, like
1240 => >ano@some.domain,ona@other.domain <original@ddress> ...
1242 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1243 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1244 not being looked up in the rDNS to get a host name, Exim was showing the
1245 IP address twice in Received: lines, even though the IP addresses were
1246 identical. For example:
1248 Received: from [1.2.3.4] (helo=[1.2.3.4])
1250 However, if the real host name was known, it was omitting the HELO data
1251 if it matched the actual IP address. This has been tidied up so that it
1252 doesn't show the same IP address twice.
1254 PH/07 When both +timestamp and +memory debugging was on, the value given by
1255 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1256 called by the debug printing, thereby overwriting the timestamp buffer.
1257 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1259 PH/08 When the original message was included in an autoreply transport, it
1260 always said "this is a copy of the message, including all the headers",
1261 even if body_only or headers_only was set. It now gives an appropriate
1264 PH/09 Applied a patch from the Sieve maintainer which:
1266 o fixes some comments
1267 o adds the (disabled) notify extension core
1268 o adds some debug output for the result of if/elsif tests
1269 o points to the current vacation draft in the documentation
1270 and documents the missing references header update
1274 o fixes a bug in processing the envelope test (when testing
1275 multiple envelope elements, the last element determinted the
1278 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1279 Electronic Mail") by including:
1281 Auto-submitted: auto-generated
1283 in the messages that it generates (bounce messages and others, such as
1284 warnings). In the case of bounce messages for non-SMTP mesages, there was
1285 also a typo: it was using "Auto_submitted" (underscore instead of
1286 hyphen). Since every message generated by Exim is necessarily in response
1287 to another message, thes have all been changed to:
1289 Auto-Submitted: auto-replied
1291 in accordance with these statements in the RFC:
1293 The auto-replied keyword:
1295 - SHOULD be used on messages sent in direct response to another
1296 message by an automatic process,
1298 - MUST NOT be used on manually-generated messages,
1300 - MAY be used on Delivery Status Notifications (DSNs) and Message
1301 Disposition Notifications (MDNs),
1303 - MUST NOT be used on messages generated by automatic or periodic
1304 processes, except for messages which are automatic responses to
1307 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1308 to the default Received: header definition.
1310 PH/12 Added log selector acl_warn_skipped (default on).
1312 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1313 variables because (a) they are in the wrong storage pool and (b) even if
1314 they were copied, it wouldn't work properly because of the caching.
1316 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1317 checking when decoding. Apparently there are clients that generate
1318 overlong encoded strings. Why am I not surprised?
1320 PH/15 If the first argument of "${if match_address" was not empty, but did not
1321 contain an "@" character, Exim crashed. Now it writes a panic log message
1322 and treats the condition as false.
1324 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1326 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1327 extension "envelope-auth". The code is finished and in agreement with
1328 other implementations, but there is no documentation so far and in fact,
1329 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1330 not changing the active code.
1332 Print executed "if" and "elsif" statements when debugging is used. This
1333 helps a great deal to understand what a filter does.
1335 Document more things not specified clearly in RFC3028. I had all this
1336 sorted out, when out of a sudden new issues came to my mind. Oops."
1338 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1341 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1342 canonical form (as documented). However, after a host name lookup from
1343 the IP address, check_host() was doing a simple string comparison with
1344 addresses acquired from the DNS when checking that the found name did
1345 have the original IP as one of its addresses. Since any found IPv6
1346 addresses are likely to be in abbreviated form, the comparison could
1347 fail. Luckily, there already exists a function for doing the comparison
1348 by converting both addresses to binary, so now that is used instead of
1349 the text comparison.
1351 PH/20 There was another similar case to PH/19, when a complete host name was
1352 given in a host list; looking up its IP address could give an abbreviated
1353 form, whereas the current host's name might or might not be abbreviated.
1354 The same fix has been applied.
1360 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1361 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1364 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1365 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1367 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1369 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1370 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1371 Also, for "Auto-Submitted", treat anything other than "no" as
1372 non-personal, in accordance with RFC 3834. (Previously it treated
1373 anything starting "auto-" as non-personal.)
1375 TF/01 The control=submission/name=... option had a problem with syntax
1376 errors if the name included a slash character. The /name= option
1377 now slurps the rest of the string, so it can include any characters
1378 but it must come last in the list of options (after /sender_retain
1381 PH/05 Some modifications to the interface to the fake nameserver for the new
1389 TK/01 Added the "success_on_redirect" address verification option. See
1390 NewStuff for rationale and an example.
1392 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1394 PH/02 Patch to exigrep to allow it to work on syslog lines.
1396 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1397 fread() to skip over the body file's header line, because in Cygwin the
1398 header line is locked and is inaccessible.
1400 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1401 co-exist for some time) to make it clear that it is the Exim ID that is
1402 referenced, not the Message-ID: header line.
1404 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1405 string_format() function, because snprintf() does not exist on all
1408 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1409 ${stat: expansion item.
1411 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1412 protocol synchronization error", to keep the pedants happy.
1414 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1415 well as for IRIX systems, when gcc is being used. See the host.c source
1418 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1420 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1423 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1424 transport and to the smtp transport in LMTP mode.
1426 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1428 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1429 run a filter in a subprocess. This could lead to confusion in subsequent
1430 lookups in the parent process. There should also be a search_tidyup() at
1431 the end of the subprocess.
1433 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1434 only if the host matched helo_try_verify_hosts, which caused the
1435 verification to occur when the EHLO/HELO command was issued. The ACL just
1436 tested the remembered result. Now, if a previous verification attempt has
1437 not happened, "verify = helo" does it there and then.
1439 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1441 TK/03 Fix log output including CR from clamd.
1443 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1444 exist provoked a memory error which could cause a segfault.
1446 PH/15 Installed PCRE 6.2
1448 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1450 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1451 of the problem. Specifically, suggested +O2 rather than +O1 for the
1454 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1456 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1457 record, the port was not correctly logged when the outgoing_port log
1458 selector was set (it logged the transort's default port).
1460 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1461 fallback_hosts, and "hosts" in the smtp transport.
1463 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1464 host errors such as "Connection refused".
1466 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1467 authentication with radiusclient 0.4.9:
1469 - Error returned from rc_read_config was caught wrongly
1470 - Username/password not passed on to radius server due to wrong length.
1472 The presumption is that some radiusclient API changes for 4.51/PH/17
1473 were not taken care of correctly. The code is still untested by me (my
1474 Linux distribution still has 0.3.2 of radiusclient), but it was
1475 contributed by a Radius user.
1477 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1478 expanding the "port" option of the smtp transport.
1480 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1481 while reading a MIME header. Thanks to Tom Hughes for a patch.
1483 PH/24 Include config.h inside local_scan.h so that configuration settings are
1486 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1487 that in an ACL for RCPT (for example), you can examine exactly what was
1490 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1491 commands, but it was not correctly comparing the address with the actual
1492 client host address. Thus, it would show the EHLO address in Received:
1493 header lines when this was not necessary.
1495 PH/27 Added the % operator to ${eval:}.
1497 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1498 it should be ignoring failures (because with -C, for example, it has lost
1499 privilege). It wasn't ignoring creation failures other than "already
1502 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1505 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1506 because an input line was too long, either on its own, or by virtue of
1507 too many continuations, the temporary file was not being removed, and the
1508 return code was incorrect.
1510 PH/31 Missing "BOOL" in function definition in filtertest.c.
1512 PH/32 Applied Sieve patches from the maintainer.
1514 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1516 PH/33 Added "verify = not_blind".
1518 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1519 Local/Makefile (with some defaults set). These are used in built scripts
1520 such as exicyclog, but they have never been used in the exim_install
1521 script (though there are many overriding facilities there). I have
1522 arranged that the exim_install script now takes note of these two
1525 PH/35 Installed configuration files for Dragonfly.
1527 PH/36 When a locally submitted message by a trusted user did not contain a
1528 From: header, and the sender address was obtained from -f or from an SMTP
1529 MAIL command, and the trusted user did not use -F to supply a sender
1530 name, $originator_name was incorrectly used when constructing a From:
1531 header. Furthermore, $originator_name was used for submission mode
1532 messages from external hosts without From: headers in a similar way,
1533 which is clearly wrong.
1535 PH/37 Added control=suppress_local_fixups.
1537 PH/38 When log_selector = +received_sender was set, and the addition of the
1538 sender made the log line's construction buffer exactly full, or one byte
1539 less than full, an overflow happened when the terminating "\n" was
1542 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1543 when the result of a list match is failure because a DNS lookup failed.
1545 PH/40 RM_COMMAND is now used in the building process.
1547 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1548 the "build-* directories that it finds.
1550 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1551 address in a domain literal was a prefix of an interface address.
1553 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1554 when verifying a sender address, unless rewrite_headers is false.
1556 PH/44 Wrote a long comment about why errors_to addresses are verified as
1557 recipients, not senders.
1559 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1560 the ratelimit ACL was added.
1562 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1564 PH/46 Added extra information about PostgreSQL errors to the error string.
1566 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1567 avoiding the need to install special zones in a real server. This is
1568 backwards compatible; if it can't find the fake resolver, it drops back.
1569 Thus, both old and new test suites can be run.
1571 TF/02 Added util/ratelimit.pl
1573 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1574 clock is set back in time.
1576 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1577 Candler <B.Candler@pobox.com>.
1579 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1580 OK for addresses that are the result of redirections.
1582 PH/48 A number of further additions for the benefit of the new test suite,
1583 including a fake gethostbyname() that interfaces to the fake DNS resolver
1586 TF/06 The fix for widen_domains has also been applied to qualify_single and
1587 search_parents which are the other dnslookup options that can cause
1590 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1593 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1595 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1596 service" if it encountered an MX record with an empty target hostname.
1597 The message is now "an MX or SRV record indicated no SMTP service".
1599 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1600 if the DNS of the sending site is misconfigured. This is quite a
1601 common situation. This change restores the behaviour of treating a
1602 helo verification defer as a failure.
1604 PH/51 If self=fail was set on a router, the bounce message did not include the
1605 actual error message.
1611 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1613 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1614 command itself ended in error, the underlying message about the transport
1615 filter timeout was being overwritten with the pipe command error. Now the
1616 underlying error message should be appended to the second error message.
1618 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1620 PH/02 Reduce the amount of output that "make" produces by default. Full output
1621 can still be requested.
1623 PH/03 The warning log line about a condition test deferring for a "warn" verb
1624 was being output only once per connection, rather than after each
1625 occurrence (because it was using the same function as for successful
1626 "warn" verbs). This seems wrong, so I have changed it.
1628 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1629 it should not have, which might have caused a crash in the right
1630 circumstances, but probably never did.
1632 PH/04 Installed a modified version of Tony Finch's patch to make submission
1633 mode fix the return path as well as the Sender: header line, and to
1634 add a /name= option so that you can make the user's friendly name appear
1637 TF/03 Added the control = fakedefer ACL modifier.
1639 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1640 Mark Lowes for thorough testing.
1642 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1644 TK/03 Merged latest SRS patch from Miles Wilton.
1646 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1647 with the definition in sysexits.h (which is #included earlier).
1648 Fortunately, Exim does not actually use EX_OK. The code used to try to
1649 preserve the sysexits.h value, by assumimg that macro definitions were
1650 scanned for macro replacements. I have been disabused of this notion,
1651 so now the code just undefines EX_OK before #including unistd.h.
1653 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1654 in the smtp transport. When a block could not be written in a single
1655 write() function, the timeout was being re-applied to each part-write.
1656 This seems wrong - if the receiver was accepting one byte at a time it
1657 would take for ever. The timeout is now adjusted when this happens. It
1658 doesn't have to be particularly precise.
1660 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1661 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1663 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1664 without a domain if the check to <postmaster@domain> fails.
1666 SC/01 Eximstats: added -xls and the ability to specify output files
1667 (patch written by Frank Heydlauf).
1669 SC/02 Eximstats: use FileHandles for outputing results.
1671 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1673 SC/04 Eximstats: fixed display of large numbers with -nvr option
1675 SC/05 Eximstats: fixed merging of reports with empty tables.
1677 SC/06 Eximstats: added the -include_original_destination flag
1679 SC/07 Eximstats: removed tabs and trailing whitespace.
1681 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1683 TK/06 MBOX spool code: Add real "From " MBOX separator line
1684 so the .eml file is really in mbox format (even though
1685 most programs do not really care). Patch from Alex Miller.
1687 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1688 The latter is generated from $received_to and is only set if the
1689 message has one envelope recipient. SA can use these headers,
1690 obviously out-of-the-box. Patch from Alex Miller.
1692 PH/08 The ${def test on a variable was returning false if the variable's
1693 value was "0", contrary to what the specification has always said!
1694 The result should be true unless the variable is empty.
1696 PH/09 The syntax error of a character other than { following "${if
1697 def:variable_name" (after optional whitespace) was not being diagnosed.
1698 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1699 accidental colon was present, for example, could give incorrect results.
1701 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1702 result is used (not including appendfile, where other changes are about
1705 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1706 This involved changing a lot of size variables from int to off_t. It
1707 should work with maildirs and everything.
1709 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1710 spamd dying while we are connected to it.
1712 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1715 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1716 function for BDB 4.3.
1718 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1720 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1721 the sources for PCRE have been re-arranged and more files are now
1724 PH/15 The code I had for printing potentially long long variables in PH/11
1725 above was not the best (it lost precision). The length of off_t variables
1726 is now inspected at build time, and an appropriate printing format (%ld
1727 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1728 to be "long long int" or "long int". This is needed for the internal
1729 formatting function string_vformat().
1731 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1732 the configuration file to be ":syslog", then the script "guesses" where
1733 the logs files are, rather than using the compiled in default. In our
1734 case the guess is not the same as the compiled default, so the script
1735 suddenly stopped working when I started to use syslog. The patch checks
1736 to see if log_file_path is "". If so, it attempts to read it from exim
1737 with no configuration file to get the compiled in version, before it
1738 falls back to the previous guessing code."
1740 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1741 implementing BATV in an Exim configuration. See NewStuff for the gory
1744 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1745 Makefile that are specific to HP-UX.
1747 PH/18 If the "use_postmaster" option was set for a recipient callout together
1748 with the "random" option, the postmaster address was used as the MAIL
1749 FROM address for the random test, but not for the subsequent recipient
1750 test. It is now used for both.
1752 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1753 patch removes a few documentation additions to RFC 3028, because the
1754 latest draft now contains them. It adds the new en;ascii-case comparator
1755 and a new error check for 8bit text in MIME parts. Comparator and
1756 require names are now matched exactly. I enabled the subaddress
1757 extension, but it is not well tested yet (read: it works for me)."
1759 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1760 rework some of the code of TK/09 above to avoid the hardwired use of
1761 "%lld" and "long long". Replaced the call to snprintf() with a call to
1764 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1765 records point to non-existent hosts", "retry timeout exceeded", and
1766 "retry time not reached for any host after a long failure period".
1768 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1769 experimental DomainKeys support:
1771 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1772 (2) On an error such as an illegally used "control", the wrong name for
1773 the control was given.
1775 These problems did NOT occur unless DomainKeys support was compiled.
1777 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1779 PH/24 Added ${if match_ip condition.
1781 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1782 databases so that it will be absolutely obvious if a crash occurs in the
1783 DB library. This is a regular occurrence (often caused by mis-matched
1786 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1787 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1788 picked up on a user's system that detects such things. There doesn't seem
1789 to be a gcc warning option for this - only an attribute that has to be
1790 put on the function's prototype. It seems that in Fedora Core 4 they have
1791 set this on a number of new functions. No doubt there will be more in due
1794 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1795 specify a transport. However, if an address that was verified by such a
1796 router was the subject of a callout, Exim crashed because it tried to
1797 read the rcpt_include_affixes from the non-existent transport. Now it
1798 just assumes that the setting of that option is false. This bug was
1799 introduced by 4.51/PH/31.
1801 PH/28 Changed -d+all to exclude +memory, because that information is very
1802 rarely of interest, but it makes the output a lot bigger. People tend to
1803 do -d+all out of habit.
1805 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1806 code in os-type was giving problems when libc.so lives in lib64, like on
1809 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1810 aren't the modern standard, and it seems that some systems' include files
1811 don't always have them. Exim was already checking for some of the newer
1812 ones like T_AAAA, and defining it itself. I've added checks for all the
1813 record types that Exim uses.
1815 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1816 not automatically generating a new one, as it is supposed to. This
1817 prevented TLS from working. If the file did exist, but contained invalid
1818 data, a new version was generated, as expected. It was only the case of a
1819 non-existent file that was broken.
1821 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1822 with a change in libdomainkeys > 0.64.
1824 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1825 from DNS. If the selector record carries the flag, it now has
1826 precedence over the domain-wide flag.
1828 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1830 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1831 the use of a port name that isn't defined in /etc/services) occurred, the
1832 message was deferred as in a normal delivery, and thus remained on the
1833 spool, instead of being failed because of the mua_wrapper setting. This
1834 is now fixed, and I tidied up some of the mua_wrapper messages at the
1837 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1838 the messages in a hash of arrays rather than using individual hashes.
1839 This is a bit cleaner and results in dramatic memory savings, albeit
1840 at a slight CPU cost.
1842 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1843 as requested by Marc Sherman.
1845 SC/10 Eximstats: added histograms for user specified patterns as requested
1848 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1850 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1851 fopen() in the content-scanning modules that did not already have it.
1857 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1858 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1860 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1862 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1864 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1865 to test Sieve filters that use "vacation".
1867 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1868 that changes the way the GnuTLS parameters are stored in the cache file.
1869 The new format can be generated externally. For backward compatibility,
1870 if the data in the cache doesn't make sense, Exim assumes it has read an
1871 old-format file, and it generates new data and writes a new file. This
1872 means that you can't go back to an older release without removing the
1875 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1876 work if there are any delivery delays because "one_time" forces the
1877 parent to be marked "delivered", so its unseen clone is never tried
1878 again. For this reason, Exim now forbids the simultaneous setting of
1881 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1882 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1883 entry if you want to know the details. The fix, however, neglected to
1884 consider the case when local delivery batching is involved. The test for
1885 "previously delivered" was not happening when checking to see if an
1886 address could be batched with a previous (undelivered) one; under
1887 certain circumstances this could lead to multiple deliveries to the same
1890 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1891 in its include files, and this causes problems building Exim.
1893 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1894 header_syntax) but Exim was just ignoring anything given after a slash.
1895 In particular, this caused confusion with an attempt to use "verify =
1896 reverse_host_lookup/defer_ok". An error is now given when options are
1897 supplied for verify items that do not have them. (Maybe reverse_host_
1898 lookup should have a defer_ok option, but that's a different point.)
1900 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1901 defined by RFC 821) to 2048, because there were problems with some AUTH
1902 commands, and RFC 1869 says the size should be increased for extended
1903 SMTP commands that take arguments.
1905 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1908 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1909 "unknown" error; now it says that the functionality isn't in the binary.
1911 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1912 an address' error message when a string expansion fails (syntax or
1913 whatever). Otherwise the password may appear in the log. Following change
1914 PH/42 below, there is no longer a chance of it appearing in a bounce
1917 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1919 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1920 of its addresses was ever tried. (Bugzilla bug #2.)
1922 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1923 the result incorrectly in the debug output. (It correctly added a newline
1924 to what was transported.)
1926 TF/01 Added $received_time.
1928 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1929 commented out examples of how to interface to a virus scanner and to
1930 SpamAssassin. Also added commented examples of av_scanner and
1931 spamd_address settings.
1933 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1934 and controls are allowed in which ACLs. There were a couple of minor
1935 errors. Some of the entries in the conditions table (which is a table of
1936 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1937 negation of where the condition IS allowed.
1939 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1941 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1942 header file does not have a version number, so I've had to invent a new
1943 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1944 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1945 radiusclient), but it was contributed by a Radius user.
1947 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1948 files or directories, for OpenSSL.
1950 PH/19 When an Exim process that is running as root has to create an Exim log
1951 file, it does so in a subprocess that runs as exim:exim so as to get the
1952 ownership right at creation (otherwise, other Exim processes might see
1953 the file with the wrong ownership). There was no test for failure of this
1954 fork() call, which would lead to the process getting stuck as it waited
1955 for a non-existent subprocess. Forks do occasionally fail when resources
1956 run out. I reviewed all the other calls to fork(); they all seem to check
1959 PH/20 When checking for unexpected SMTP input at connect time (before writing
1960 the banner), Exim was not dealing correctly with a non-positive return
1961 from the read() function. If the client had disconnected by this time,
1962 the result was a log entry for a synchronization error with an empty
1963 string after "input=" when read() returned zero. If read() returned -1
1964 (an event I could not check), uninitialized data bytes were printed.
1965 There were reports of junk text (parts of files, etc) appearing after
1968 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1970 PH/22 Added support for macro redefinition, and (re)definition in between
1971 driver and ACL definitions.
1973 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1974 forgetting to use the resulting value; it was using the unexpanded value.
1976 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1977 hadn't been configured. The fix is from Juergen Kreileder, who
1978 understands it better than I do:
1980 "Here's what I see happening with three configured cyrus_sasl
1981 authenticators configured (plain, login, cram-md5):
1983 On startup auth_cyrus_sasl_init() gets called for each of these.
1984 This means three calls to sasl_listmech() without a specified mech_list.
1985 => SASL tests which mechs of all available mechs actually work
1986 => three warnings about OTP not working
1987 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1989 With the patch, sasl_listmech() also gets called three times. But now
1990 SASL's mech_list option is set to the server_mech specified in the the
1991 authenticator. Or in other words, the answer from sasl_listmech()
1992 gets limited to just the mech you're testing for (which is different
1994 => the return list contains just 'plain' or 'login', 'cram-md5' or
1995 nothing depending on the value of ob->server_mech.
1997 I've just tested the patch: Authentication still works fine,
1998 unavailable mechs specified in the exim configuration are still
1999 caught, and the auth.log warnings about OTP are gone."
2001 PH/25 When debugging is enabled, the contents of the command line are added
2002 to the debugging output, even when log_selector=+arguments is not
2005 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2006 answer is "GNU", and only if the return is "GNU/something" is the answer
2009 PH/27 $acl_verify_message is now set immediately after the failure of a
2010 verification in an ACL, and so is available in subsequent modifiers. In
2011 particular, the message can be preserved by coding like this:
2013 warn !verify = sender
2014 set acl_m0 = $acl_verify_message
2016 Previously, $acl_verify_message was set only while expanding "message"
2017 and "log_message" when a very denied access.
2019 PH/28 Modified OS/os.c-Linux with
2021 -#ifndef OS_LOAD_AVERAGE
2022 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2024 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2025 nomenclature these days.)
2027 PH/29 Installed patch from the Sieve maintainer that adds the options
2028 sieve_useraddress and sieve_subaddress to the redirect router.
2030 PH/30 In these circumstances:
2031 . Two addresses routed to the same list of hosts;
2032 . First host does not offer TLS;
2033 . First host accepts first address;
2034 . First host gives temporary error to second address;
2035 . Second host offers TLS and a TLS session is established;
2036 . Second host accepts second address.
2037 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2038 and peerdn, if requested) that were in fact used only for the second
2041 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2042 attention to any local part prefix or suffix that was matched by the
2043 router that accepted the address. It now behaves in the same way as it
2044 does for delivery: the affixes are removed from the local part unless
2045 rcpt_include_affixes is set on the transport.
2047 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2048 timeout during the DATA phase of an incoming message.
2050 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2051 applied a patch sanctioned by the Sieve maintainer.
2053 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2054 the uid or gid is negative. A case of a negative gid caused this to be
2055 noticed. The fix allows for either to be negative.
2057 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2058 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2061 PH/36 The change PH/12 above was broken. Fixed it.
2063 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2064 the grounds that routing the same address twice would always produce the
2065 same answer. This might have been true once, but it is certainly no
2066 longer true now. Routing a child address may depend on the previous
2067 routing that produced that child. Some complicated redirection strategies
2068 went wrong when messages had multiple recipients, and made Exim's
2069 behaviour dependent on the order in which the addresses were given.
2071 I have moved the duplicate checking until after the routing is complete.
2072 Exim scans the addresses that are assigned to local and remote
2073 transports, and removes any duplicates. This means that more work will be
2074 done, as duplicates will always all be routed, but duplicates are
2075 presumably rare, so I don't expect this is of any significance.
2077 For deliveries to pipes, files, and autoreplies, the duplicate checking
2078 still happens during the routing process, since they are not going to be
2081 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2082 It corrects a timeout issue with spamd. This is Ian's comment: "The
2083 background is that sometimes spamd either never reads data from a
2084 connection it has accepted, or it never writes response data. The exiscan
2085 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2086 blindly assumes that writes won't block so it may never time out."
2088 PH/39 Allow G after quota size as well as K and M.
2090 PH/40 The value set for $authenticated_id in an authenticator may not contain
2091 binary zeroes or newlines because the value is written to log lines and
2092 to spool files. There was no check on this. Now the value is run through
2093 the string_printing() function so that such characters are converted to
2094 printable escape sequences.
2096 PH/41 $message_linecount is a new variable that contains the total number of
2097 lines in the message. Compare $body_linecount, which is the count for the
2100 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2101 bounce and delay warning messages, except in certain special cases, which
2104 (a) An SMTP error message from a remote host;
2105 (b) A message specified in a :fail: redirection;
2106 (c) A message specified in a "fail" command in a system filter;
2107 (d) A message specified in a FAIL return from the queryprogram router;
2108 (e) A message specified by the cannot_route_message router option.
2110 In these cases only, Exim does include the error details in bounce and
2111 warning messages. There are also a few cases where bland messages such
2112 as "unrouteable address" or "local delivery error" are given.
2114 PH/43 $value is now also set for the "else" part of a ${run expansion.
2116 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2117 being worked on, but at least Exim now implements the latest version to
2120 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2121 process to complete was treated as a delivery failure, a timeout while
2122 writing the message to the pipe was logged, but erroneously treated as a
2123 successful delivery. Such timeouts include transport filter timeouts. For
2124 consistency with the overall process timeout, these timeouts are now
2125 treated as errors, giving rise to delivery failures by default. However,
2126 there is now a new Boolean option for the pipe transport called
2127 timeout_defer, which, if set TRUE, converts the failures into defers for
2128 both kinds of timeout. A transport filter timeout is now identified in
2131 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2132 systems where "make" and "gmake" are different, calling "gmake" at top
2133 level broke things. I've arranged for the value of $(MAKE) to be passed
2134 from the Makefile to this script so that it can call the same version of
2138 A note about Exim versions 4.44 and 4.50
2139 ----------------------------------------
2141 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2142 changes of various kinds. As a consequence, a big documentation update was
2143 needed. This delayed the release for rather longer than seemed good, especially
2144 in the light of a couple of (minor) security issues. Therefore, the changes
2145 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2146 release. So 4.44 and 4.50 are in effect two different branches that both start
2149 I have left the 4.50 change log unchanged; it contains all the changes since
2150 4.43. The change log for 4.44 is below; many of its items are identical to
2151 those for 4.50. This seems to be the most sensible way to preserve the
2152 historical information.
2158 1. Minor wording change to the doc/README.SIEVE file.
2160 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2161 computation of the current number of files was incorrect.
2163 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2164 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2167 4. Give more explanation in the error message when the command for a transport
2168 filter fails to execute.
2170 5. There are several places where Exim runs a non-Exim command in a
2171 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2172 was being done only for the command run by the queryprogram router. It is
2173 now done for all such subprocesses. The other cases are: ${run, transport
2174 filters, and the commands run by the lmtp and pipe transports.
2176 6. Added CONFIGURE_GROUP build-time option.
2178 7. Some older OS have a limit of 256 on the maximum number of file
2179 descriptors. Exim was using setrlimit() to set 1000 as a large value
2180 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2181 systems. I've change it so that if it can't get 1000, it tries for 256.
2183 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2184 was an oversight, and furthermore, ever since the addition of extra
2185 controls (e.g. 4.43/32), the checks on when to allow different forms of
2186 "control" were broken. There should now be diagnostics for all cases when a
2187 control that does not make sense is encountered.
2189 9. Added the /retain_sender option to "control=submission".
2191 10. $recipients is now available in the predata ACL (oversight).
2193 11. Tidy the search cache before the fork to do a delivery from a message
2194 received from the command line. Otherwise the child will trigger a lookup
2195 failure and thereby defer the delivery if it tries to use (for example) a
2196 cached ldap connection that the parent has called unbind on.
2198 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2199 of $address_data from the recipient verification was clobbered by the
2200 sender verification.
2202 13. The value of address_data from a sender verification is now available in
2203 $sender_address_data in subsequent conditions in the ACL statement.
2205 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2207 15. Added a new option "connect=<time>" to callout options, to set a different
2210 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2211 was its contents. (It was OK if the option was not defined at all.)
2213 17. A "Completed" log line is now written for messages that are removed from
2214 the spool by the -Mrm option.
2216 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2217 information about exactly what failed.
2219 19. Added -dd to debug only the daemon process.
2221 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2222 handles timeouts, both on the server side and network timeouts. Renamed the
2223 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2225 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2226 It is now set to "smtps".
2228 22. $host_address is now set to the target address during the checking of
2229 ignore_target_hosts.
2231 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2232 being passed; this would have caused $sender_host_name to have been used if
2233 matching the list had actually called for a host name (not very likely,
2234 since this list is usually IP addresses). A host name is now passed as
2237 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2238 code that specifies a non-restarting handler (typically sigaction() in
2239 modern systems) in an attempt to fix a rare and obscure crash bug.
2241 25. Narrowed the window for a race in the daemon that could cause it to ignore
2242 SIGCHLD signals. This is not a major problem, because they are used only to
2243 wake it up if nothing else does.
2245 26. A malformed maildirsize file could cause Exim to calculate negative values
2246 for the mailbox size or file count. Odd effects could occur as a result.
2247 The maildirsize information is now recalculated if the size or filecount
2250 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2251 support for a long time. Removed HAVE_SYS_VFS_H.
2253 28. Installed the latest version of exipick from John Jetmore.
2255 29. In an address list, if the pattern was not a regular expression, an empty
2256 subject address (from a bounce message) matched only if the pattern was an
2257 empty string. Non-empty patterns were not even tested. This was the wrong
2258 because it is perfectly reasonable to use an empty address as part of a
2259 database query. An empty address is now tested by patterns that are
2260 lookups. However, all the other forms of pattern expect the subject to
2261 contain a local part and a domain, and therefore, for them, an empty
2262 address still always fails if the pattern is not itself empty.
2264 30. Exim went into a mad DNS loop when attempting to do a callout where the
2265 host was specified on an smtp transport, and looking it up yielded more
2266 than one IP address.
2268 31. Re-factored the code for checking spool and log partition space into a
2269 function that finds that data and another that does the check. The former
2270 is then used to implement four new variables: $spool_space, $log_space,
2271 $spool_inodes, and $log_inodes.
2273 32. The RFC2047 encoding function was originally intended for short strings
2274 such as real names; it was not keeping to the 75-character limit for
2275 encoded words that the RFC imposes. It now respects the limit, and
2276 generates multiple encoded words if necessary. To be on the safe side, I
2277 have increased the buffer size for the ${rfc2047: expansion operator from
2280 33. It is now permitted to omit both strings after an "if" condition; if the
2281 condition is true, the result is "true". As before, when the second string
2282 is omitted, a false condition yields an empty string. This makes it less
2283 cumbersome to write custom ACL and router conditions.
2285 34. Failure to deliver a bounce message always caused it to be frozen, even if
2286 there was an errors_to setting on the router. The errors_to setting is now
2289 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2290 canonical form (fully expanded) before being placed in
2291 $sender_host_address.
2293 36. The table in the code that translates DNS record types into text (T_A to
2294 "A" for instance) was missing entries for NS and CNAME. It is just possible
2295 that this could have caused confusion if both these types were looked up
2296 for the same domain, because the text type is used as part of Exim's
2297 per-process caching. But the chance of anyone hitting this buglet seems
2300 37. The dnsdb lookup has been extended in a number of ways.
2302 (1) There is a new type, "zns", which walks up the domain tree until it
2303 finds some nameserver records. It should be used with care.
2305 (2) There is a new type, "mxh", which is like "mx" except that it returns
2306 just the host names, not the priorities.
2308 (3) It is now possible to give a list of domains (or IP addresses) to be
2309 looked up. The behaviour when one of the lookups defers can be
2310 controlled by a keyword.
2312 (4) It is now possible to specify the separator character for use when
2313 multiple records are returned.
2315 38. The dnslists ACL condition has been extended: it is now possible to supply
2316 a list of IP addresses and/or domains to be looked up in a particular DNS
2319 39. Added log_selector=+queue_time_overall.
2321 40. When running the queue in the test harness, wait just a tad after forking a
2322 delivery process, to get repeatability of debugging output.
2324 41. Include certificate and key file names in error message when GnuTLS fails
2325 to set them up, because the GnuTLS error message doesn't include the name
2326 of the failing file when there is a problem reading it.
2328 42. Allow both -bf and -bF in the same test run.
2330 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2332 44. The "Exiscan patch" is now merged into the mainline Exim source.
2334 45. Sometimes the final signoff response after QUIT could fail to get
2335 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2336 < 0 before doing a fflush(). This bug looks as though it goes back to the
2337 introduction of TLS in release 3.20, but "sometimes" must have been rare
2338 because the tests only now provoked it.
2340 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2341 (this can affect the format of dates).
2343 47. exim_tidydb, when checking for the continued existence of a message for
2344 which it has found a message-specific retry record, was not finding
2345 messages that were in split spool directories. Consequently, it was
2346 deleting retry records that should have stayed in existence.
2348 48. Steve fixed some bugs in eximstats.
2350 49. The SPA authentication driver was not abandoning authentication and moving
2351 on to the next authenticator when an expansion was forced to fail,
2352 contradicting the general specification for all authenticators. Instead it
2353 was generating a temporary error. It now behaves as specified.
2355 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2356 (the order specifies the preference for clients). The order is now AES256,
2357 AES128, 3DES, ARCFOUR128.
2359 51. Small patch to Sieve code - explicitly set From: when generating an
2362 52. Exim crashed if a remote delivery caused a very long error message to be
2363 recorded - for instance if somebody sent an entire SpamAssassin report back
2364 as a large number of 550 error lines. This bug was coincidentally fixed by
2365 increasing the size of one of Exim's internal buffers (big_buffer) that
2366 happened as part of the Exiscan merge. However, to be on the safe side, I
2367 have made the code more robust (and fixed the comments that describe what
2370 53. Now that there can be additional text after "Completed" in log lines (if
2371 the queue_time_overall log selector is set), a one-byte patch to exigrep
2372 was needed to allow it to recognize "Completed" as not the last thing in
2375 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2376 patch that reportedly fixes this has been added. I am not expert enough to
2377 create a test for it. This is what the patch creator wrote:
2379 "I found a little strange behaviour of ldap code when working with
2380 Windows 2003 AD Domain, where users was placed in more than one
2381 Organization Units. When I tried to give exim partial DN, the exit code
2382 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2383 But simultaneously result of request was absolutely normal ldap result,
2384 so I produce this patch..."
2386 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2387 REFERENCE, so I have modified the code to exclude the patch when that macro
2390 55. Some experimental protocols are using DNS PTR records for new purposes. The
2391 keys for these records are domain names, not reversed IP addresses. The
2392 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2393 leaves it alone. Component reversal etc. now happens only for IP addresses.
2396 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2398 57. Double the size of the debug message buffer (to 2048) so that more of very
2399 long debug lines gets shown.
2401 58. The exicyclog utility now does better if the number of log files to keep
2402 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2404 59. Two changes related to the smtp_active_hostname option:
2406 (1) $smtp_active_hostname is now available as a variable.
2407 (2) The default for smtp_banner uses $smtp_active_hostname instead
2408 of $primary_hostname.
2410 60. The host_aton() function is supposed to be passed a string that is known
2411 to be a valid IP address. However, in the case of IPv6 addresses, it was
2412 not checking this. This is a hostage to fortune. Exim now panics and dies
2413 if the condition is not met. A case was found where this could be provoked
2414 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2415 components; fortuitously, this particular loophole had already been fixed
2416 by change 4.50/55 above.
2418 If there are any other similar loopholes, the new check in host_aton()
2419 itself should stop them being exploited. The report I received stated that
2420 data on the command line could provoke the exploit when Exim was running as
2421 exim, but did not say which command line option was involved. All I could
2422 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2423 running as the user.
2426 61. There was a buffer overflow vulnerability in the SPA authentication code
2427 (which came originally from the Samba project). I have added a test to the
2428 spa_base64_to_bits() function which I hope fixes it.
2431 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2432 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2434 63. The daemon start-up calls getloadavg() while still root for those OS that
2435 need the first call to be done as root, but it missed one case: when
2436 deliver_queue_load_max is set with deliver_drop_privilege. This is
2437 necessary for the benefit of the queue runner, because there is no re-exec
2438 when deliver_drop_privilege is set.
2440 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2441 This has been fixed.
2443 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2444 was in use, was not putting the data itself into the right store pool;
2445 consequently, it could be overwritten for a subsequent message in the same
2446 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2449 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2451 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2452 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2453 the function were treating the return as a boolean value, which happened to
2454 work because 0=false and not-0=true, but is not correct code.
2456 68. The host_aton() function was not handling scoped IPv6 addresses (those
2457 with, for example, "%eth0" on the end) correctly.
2459 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2460 negated items (that is, ~something) in unsigned ints. Some compilers
2461 apparently mutter when there is no cast.
2463 70. If an address verification called from an ACL failed, and did not produce a
2464 user-specific message (i.e. there was only a "system" message), nothing was
2465 put in $acl_verify_message. In this situation, it now puts the system
2468 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2469 session; change 4.31/43 added the unwanted input to the log line - except
2470 that it did not do this in the start of session case. It now does.
2472 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2473 This is wrong and can cause the other end to generate a synchronization
2474 error if it is another Exim or anything else that does the synchronization
2475 check. A QUIT command is no longer sent after a timeout.
2477 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2478 during host lookups.
2480 74. The defer_ok option of callout verification was not working if it was used
2481 when verifying addresses in header lines, that is, for this case:
2483 verify = header_sender/callout=defer_ok
2485 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2486 those file descriptors could be used for SMTP connections. If anything
2487 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2488 could be sent to the SMTP client, causing chaos. The daemon now opens
2489 stdin, stdout, and stderr to /dev/null when it puts itself into the
2492 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2493 log by default. The user can override this with suitable Perl magic.
2495 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2496 the log message when discard triggers, was not working for the DATA ACL or
2497 for the non-SMTP ACL.
2499 78. Error message wording change in sieve.c.
2501 79. If smtp_accept_max_per_host was set, the number of connections could be
2502 restricted to fewer than expected, because the daemon was trying to set up
2503 a new connection before checking whether the processes handling previous
2504 connections had finished. The check for completed processes is now done
2505 earlier. On busy systems, this bug wouldn't be noticed because something
2506 else would have woken the daemon, and it would have reaped the completed
2509 80. If a message was submitted locally by a user whose login name contained one
2510 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2511 It caused a spool format error. I have fixed the spool reading code. A
2512 related problem was that the "from" clause in the Received: line became
2513 illegal because of the space(s). It is now covered by ${quote_local_part.
2515 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2518 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2520 83. Patch from Sieve maintainer to fix unterminated string problem in
2521 "vacation" handling.
2523 84. Some minor changes to the Linux configuration files to help with other
2524 OS variants using glibc.
2526 85. One more patch for Sieve to update vacation handling to latest spec.
2529 ----------------------------------------------------
2530 See the note above about the 4.44 and 4.50 releases.
2531 ----------------------------------------------------
2537 1. Change 4.43/35 introduced a bug that caused file counts to be
2538 incorrectly computed when quota_filecount was set in an appendfile
2541 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2542 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2545 3. Give more explanation in the error message when the command for a transport
2546 filter fails to execute.
2548 4. There are several places where Exim runs a non-Exim command in a
2549 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2550 was being done only for the command run by the queryprogram router. It is
2551 now done for all such subprocesses. The other cases are: ${run, transport
2552 filters, and the commands run by the lmtp and pipe transports.
2554 5. Some older OS have a limit of 256 on the maximum number of file
2555 descriptors. Exim was using setrlimit() to set 1000 as a large value
2556 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2557 systems. I've change it so that if it can't get 1000, it tries for 256.
2559 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2560 was an oversight, and furthermore, ever since the addition of extra
2561 controls (e.g. 4.43/32), the checks on when to allow different forms of
2562 "control" were broken. There should now be diagnostics for all cases when a
2563 control that does not make sense is encountered.
2565 7. $recipients is now available in the predata ACL (oversight).
2567 8. Tidy the search cache before the fork to do a delivery from a message
2568 received from the command line. Otherwise the child will trigger a lookup
2569 failure and thereby defer the delivery if it tries to use (for example) a
2570 cached ldap connection that the parent has called unbind on.
2572 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2573 of $address_data from the recipient verification was clobbered by the
2574 sender verification.
2576 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2577 was its contents. (It was OK if the option was not defined at all.)
2579 11. A "Completed" log line is now written for messages that are removed from
2580 the spool by the -Mrm option.
2582 12. $host_address is now set to the target address during the checking of
2583 ignore_target_hosts.
2585 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2586 being passed; this would have caused $sender_host_name to have been used if
2587 matching the list had actually called for a host name (not very likely,
2588 since this list is usually IP addresses). A host name is now passed as
2591 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2592 code that specifies a non-restarting handler (typically sigaction() in
2593 modern systems) in an attempt to fix a rare and obscure crash bug.
2595 15. Narrowed the window for a race in the daemon that could cause it to ignore
2596 SIGCHLD signals. This is not a major problem, because they are used only to
2597 wake it up if nothing else does.
2599 16. A malformed maildirsize file could cause Exim to calculate negative values
2600 for the mailbox size or file count. Odd effects could occur as a result.
2601 The maildirsize information is now recalculated if the size or filecount
2604 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2605 support for a long time. Removed HAVE_SYS_VFS_H.
2607 18. Updated exipick to current release from John Jetmore.
2609 19. Allow an empty sender to be matched against a lookup in an address list.
2610 Previously the only cases considered were a regular expression, or an
2613 20. Exim went into a mad DNS lookup loop when doing a callout where the
2614 host was specified on the transport, if the DNS lookup yielded more than
2617 21. The RFC2047 encoding function was originally intended for short strings
2618 such as real names; it was not keeping to the 75-character limit for
2619 encoded words that the RFC imposes. It now respects the limit, and
2620 generates multiple encoded words if necessary. To be on the safe side, I
2621 have increased the buffer size for the ${rfc2047: expansion operator from
2624 22. Failure to deliver a bounce message always caused it to be frozen, even if
2625 there was an errors_to setting on the router. The errors_to setting is now
2628 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2629 canonical form (fully expanded) before being placed in
2630 $sender_host_address.
2632 24. Updated eximstats to version 1.33
2634 25. Include certificate and key file names in error message when GnuTLS fails
2635 to set them up, because the GnuTLS error message doesn't include the name
2636 of the failing file when there is a problem reading it.
2638 26. Expand error message when OpenSSL has problems setting up cert/key files.
2641 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2642 (this can affect the format of dates).
2644 28. exim_tidydb, when checking for the continued existence of a message for
2645 which it has found a message-specific retry record, was not finding
2646 messages that were in split spool directories. Consequently, it was
2647 deleting retry records that should have stayed in existence.
2649 29. eximstats updated to version 1.35
2650 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2651 1.35 - bugfix such that pie charts by volume are generated correctly
2653 30. The SPA authentication driver was not abandoning authentication and moving
2654 on to the next authenticator when an expansion was forced to fail,
2655 contradicting the general specification for all authenticators. Instead it
2656 was generating a temporary error. It now behaves as specified.
2658 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2659 (the order specifies the preference for clients). The order is now AES256,
2660 AES128, 3DES, ARCFOUR128.
2662 31. Small patch to Sieve code - explicitly set From: when generating an
2665 32. Exim crashed if a remote delivery caused a very long error message to be
2666 recorded - for instance if somebody sent an entire SpamAssassin report back
2667 as a large number of 550 error lines. This bug was coincidentally fixed by
2668 increasing the size of one of Exim's internal buffers (big_buffer) that
2669 happened as part of the Exiscan merge. However, to be on the safe side, I
2670 have made the code more robust (and fixed the comments that describe what
2673 33. Some experimental protocols are using DNS PTR records for new purposes. The
2674 keys for these records are domain names, not reversed IP addresses. The
2675 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2676 leaves it alone. Component reversal etc. now happens only for IP addresses.
2679 34. The host_aton() function is supposed to be passed a string that is known
2680 to be a valid IP address. However, in the case of IPv6 addresses, it was
2681 not checking this. This is a hostage to fortune. Exim now panics and dies
2682 if the condition is not met. A case was found where this could be provoked
2683 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2684 components; fortuitously, this particular loophole had already been fixed
2685 by change 4.50/55 or 4.44/33 above.
2687 If there are any other similar loopholes, the new check in host_aton()
2688 itself should stop them being exploited. The report I received stated that
2689 data on the command line could provoke the exploit when Exim was running as
2690 exim, but did not say which command line option was involved. All I could
2691 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2692 running as the user.
2695 35. There was a buffer overflow vulnerability in the SPA authentication code
2696 (which came originally from the Samba project). I have added a test to the
2697 spa_base64_to_bits() function which I hope fixes it.
2700 36. The daemon start-up calls getloadavg() while still root for those OS that
2701 need the first call to be done as root, but it missed one case: when
2702 deliver_queue_load_max is set with deliver_drop_privilege. This is
2703 necessary for the benefit of the queue runner, because there is no re-exec
2704 when deliver_drop_privilege is set.
2706 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2707 was in use, was not putting the data itself into the right store pool;
2708 consequently, it could be overwritten for a subsequent message in the same
2709 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2712 38. Sometimes the final signoff response after QUIT could fail to get
2713 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2714 < 0 before doing a fflush(). This bug looks as though it goes back to the
2715 introduction of TLS in release 3.20, but "sometimes" must have been rare
2716 because the tests only now provoked it.
2722 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2723 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2724 Now it returns an integer. A number of calls were still expecting a T/F
2725 return. Fortuitously, in all cases, the tests worked in OK situations,
2726 which is the norm. However, things would have gone wrong on any write
2727 failures on the smtp file descriptor. This function is used when sending
2728 messages over SMTP and also when doing verify callouts.
2730 2. When Exim is called to do synchronous delivery of a locally submitted
2731 message (the -odf or -odi options), it no longer closes stderr before doing
2734 3. Implemented the mua_wrapper option.
2736 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2738 5. Implemented the functions header_remove(), header_testname(),
2739 header_add_at_position(), and receive_remove_recipient(), and exported them
2742 6. If an ACL "warn" statement specified the addition of headers, Exim already
2743 inserted X-ACL-Warn: at the start if there was no header name. However, it
2744 was not making this test for the second and subsequent header lines if
2745 there were newlines in the string. This meant that an invalid header could
2746 be inserted if Exim was badly configured.
2748 7. Allow an ACL "warn" statement to add header lines at the start or after all
2749 the Received: headers, as well as at the end.
2751 8. Added the rcpt_4xx retry error code.
2753 9. Added postmaster_mailfrom=xxx to callout verification option.
2755 10. Added mailfrom=xxxx to the callout verification option, for verify=
2758 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2759 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2761 12. Inserted some casts to stop certain compilers warning when using pointer
2762 differences as field lengths or precisions in printf-type calls (mostly
2763 affecting debugging statements).
2765 13. Added optional readline() support for -be (dynamically loaded).
2767 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2768 same clock tick as a message's arrival, so that its received time was the
2769 same as the "first fail" time on the retry record, and that message
2770 remained on the queue past the ultimate address timeout, every queue runner
2771 would try a delivery (because it was past the ultimate address timeout) but
2772 after another failure, the ultimate address timeout, which should have then
2773 bounced the address, did not kick in. This was a "< instead of <=" error;
2774 in most cases the first failure would have been in the next clock tick
2775 after the received time, and all would be well.
2777 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2778 being recognized when the domain list was tested by the match_domain
2779 condition in an expansion string.
2781 16. Added the ${str2b64: operator.
2783 17. Exim was always calling setrlimit() to set a large limit for the number of
2784 processes, without checking whether the existing limit was already
2785 adequate. (It did check for the limit on file descriptors.) Furthermore,
2786 errors from getrlimit() and setrlimit() were being ignored. Now they are
2787 logged to the main and panic logs, but Exim does carry on, to try to do its
2788 job under whatever limits there are.
2790 18. Imported PCRE 5.0.
2792 19. Trivial typo in log message " temporarily refused connection" (the leading
2795 20. If the log selector return_path_on_delivery was set and an address was
2796 redirected to /dev/null, the delivery process crashed because it assumed
2797 that a return path would always be set for a "successful" delivery. In this
2798 case, the whole delivery is bypassed as an optimization, and therefore no
2801 21. Internal re-arrangement: the function for sending a challenge and reading
2802 a response while authentication was assuming a zero-terminated challenge
2803 string. It's now changed to take a pointer and a length, to allow for
2804 binary data in such strings.
2806 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2808 23. Exim was not respecting finduser_retries when seeking the login of the
2809 uid under which it was called; it was always trying 10 times. (The default
2810 setting of finduser_retries is zero.) Also, it was sleeping after the final
2811 failure, which is pointless.
2813 24. Implemented tls_on_connect_ports.
2815 25. Implemented acl_smtp_predata.
2817 26. If the domain in control=submission is set empty, Exim assumes that the
2818 authenticated id is a complete email address when it generates From: or
2819 Sender: header lines.
2821 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2822 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2823 chown and chgrp in /bin and hostname in /usr/bin.
2825 28. Exim was keeping the "process log" file open after each use, just as it
2826 does for the main log. This opens the possibility of it remaining open for
2827 long periods when the USR1 signal hits a daemon. Occasional processlog
2828 errors were reported, that could have been caused by this. Anyway, it seems
2829 much more sensible not to leave this file open at all, so that is what now
2832 29. The long-running daemon process does not normally write to the log once it
2833 has entered its main loop, and it closes the log before doing so. This is
2834 so that log files can straightforwardly be renamed and moved. However,
2835 there are a couple of unusual error situations where the daemon does write
2836 log entries, and I had neglected to close the log afterwards.
2838 30. The text of an SMTP error response that was received during a remote
2839 delivery was being truncated at 512 bytes. This is too short for some of
2840 the long messages that one sometimes sees. I've increased the limit to
2843 31. It is now possible to make retry rules that apply only when a message has a
2844 specific sender, in particular, an empty sender.
2846 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2847 it possible to be selective about when SMTP synchronization is enforced.
2849 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2851 32. Implemented hosts_connection_nolog.
2853 33. Added an ACL for QUIT.
2855 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2858 35. Added mailbox_size and mailbox_filecount to appendfile.
2860 36. Added control = no_multiline_responses to ACLs.
2862 37. There was a bug in the logic of the code that waits for the clock to tick
2863 in the case where the clock went backwards by a substantial amount such
2864 that the microsecond fraction of "now" was more than the microsecond
2865 fraction of "then" (but the whole seconds number was less).
2867 38. Added support for the libradius Radius client library this is found on
2868 FreeBSD (previously only the radiusclient library was supported).
2874 1. When certain lookups returned multiple values in the form name=value, the
2875 quoting of the values was not always being done properly. Specifically:
2876 (a) If the value started with a double quote, but contained no whitespace,
2878 (b) If the value contained whitespace other than a space character (i.e.
2879 tabs or newlines or carriage returns) it was not quoted.
2880 This fix has been applied to the mysql and pgsql lookups by writing a
2881 separate quoting function and calling it from the lookup code. The fix
2882 should probably also be applied to nisplus, ibase and oracle lookups, but
2883 since I cannot test any of those, I have not disturbed their existing code.
2885 2. A hit in the callout cache for a specific address caused a log line with no
2886 reason for rejecting RCPT. Now it says "Previous (cached) callout
2887 verification failure".
2889 3. There was an off-by-one bug in the queryprogram router. An over-long
2890 return line was truncated at 256 instead of 255 characters, thereby
2891 overflowing its buffer with the terminating zero. As well as fixing this, I
2892 have increased the buffer size to 1024 (and made a note to document this).
2894 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2895 when Exim is waiting for an SMTP response from a remote server, Exim
2896 restarts its select() call on the socket, thereby resetting its timeout.
2897 This is not a problem when such interrupts are rare. Somebody set up a cron
2898 job to run exiwhat every 2 minutes, which is less than the normal select()
2899 timeout (5 or 10 minutes). This meant that the select() timeout never
2900 kicked in because it was always reset. I have fixed this by comparing the
2901 time when an interrupt arrives with the time at the start of the first call
2902 to select(). If more time than the timeout has elapsed, the interrupt is
2903 treated as a timeout.
2905 5. Some internal re-factoring in preparation for the addition of Sieve
2906 extensions (by MH). In particular, the "personal" test is moved to a
2907 separate function, and given an option for scanning Cc: and Bcc: (which is
2908 not set for Exim filters).
2910 6. When Exim created an email address using the login of the caller as the
2911 local part (e.g. when creating a From: or Sender: header line), it was not
2912 quoting the local part when it contained special characters such as @.
2914 7. Installed new OpenBSD configuration files.
2916 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2917 try to make them clearer.
2919 9. Callout options, other than the timeout value, were being ignored when
2920 verifying sender addresses in header lines. For example, when using
2922 verify = header_sender/callout=no_cache
2924 the cache was (incorrectly) being used.
2926 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2927 only the Cygwin environment.
2929 11. When return_path_on_delivery was set as a log selector, if different remote
2930 addresses in the same message used different return paths and parallel
2931 remote delivery occurred, the wrong values would sometimes be logged.
2932 (Whenever a remote delivery process finished, the return path value from
2933 the most recently started remote delivery process was logged.)
2935 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2936 lines when AUTH and/or TLS are in use. This is the "received protocol"
2937 field. Exim used to use "asmtp" for authenticated SMTP, without any
2938 indication (in the protocol name) for TLS use. Now it follows the RFC and
2939 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2940 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2941 names appear in log lines as well as in Received: header lines.
2943 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2944 extensions, and comparison tests, and to fix some bugs.
2946 14. Changes to the "personal" filter test:
2948 (1) The test was buggy in that it was just doing the equivalent of
2949 "contains" tests on header lines. For example, if a user's address was
2950 anne@some.where, the "personal" test would incorrectly be true for
2952 To: susanne@some.where
2954 This test is now done by extracting each address from the header in turn,
2955 and checking the entire address. Other tests that are part of "personal"
2956 are now done using regular expressions (for example, to check local parts
2957 of addresses in From: header lines).
2959 (2) The list of non-personal local parts in From: addresses has been
2960 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2961 taken from the Sieve specification recommendations.
2963 (3) If the message contains any header line starting with "List-" it is
2964 treated as non-personal.
2966 (4) The test for "circular" in the Subject: header line has been removed
2967 because it now seems ill-conceived.
2969 15. Minor typos in src/EDITME comments corrected.
2971 16. Installed latest exipick from John Jetmore.
2973 17. If headers_add on a router specified a text string that was too long for
2974 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2975 of string_sprintf() is now avoided.
2977 18. $message_body_size was not set (it was always zero) when running the DATA
2978 ACL and the local_scan() function.
2980 19. For the "mail" command in an Exim filter, no default was being set for
2981 the once_repeat time, causing a random time value to be used if "once" was
2982 specified. (If the value happened to be <= 0, no repeat happened.) The
2983 default is now 0s, meaning "never repeat". The "vacation" command was OK
2984 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2985 (I found it when inspecting the code).
2987 20. There is now an overall timeout for performing a callout verification. It
2988 defaults to 4 times the callout timeout, which applies to individual SMTP
2989 commands during the callout. The overall timeout applies when there is more
2990 than one host that can be tried. The timeout is checked before trying the
2991 next host. This prevents very long delays if there are a large number of
2992 hosts and all are timing out (e.g. when the network connections are timing
2993 out). The value of the overall timeout can be changed by specifying an
2994 additional sub-option for "callout", called "maxwait". For example:
2996 verify = sender/callout=5s,maxwait=20s
2998 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2999 to the end before writing, but this should make it even safer).
3001 22. Exim was forgetting that it had advertised PIPELINING for the second and
3002 subsequent messages on an SMTP connection. It was also not resetting its
3003 memory on STARTTLS and an internal HELO.
3005 23. When Exim logs an SMTP synchronization error within a session, it now
3006 records whether PIPELINING has been advertised or not.
3008 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3009 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3010 rather than long int.
3012 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3014 26. Added the never_mail option to autoreply.
3020 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3021 crash if the getsockname() call failed; this can happen if a connection is
3022 closed very soon after it is established. The problem was simply in the
3023 order in which certain operations were done, causing Exim to try to write
3024 to the SMTP stream before it had set up the file descriptor. The bug has
3025 been fixed by making things happen in the correct order.
3031 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3032 before the connection was closed, thus losing the rejection response.
3034 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3035 some early Solaris releases, but causes trouble in current releases where
3036 socklen_t is defined.
3038 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3041 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3044 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3045 that releasing the top of it at the end releases what was used for sub-
3046 expansions (unless the block got too big). However, discard this block if
3047 the first thing is a variable or header, so that we can use its block when
3048 it is dynamic (useful for very large $message_headers, for example).
3050 6. Lookups now cache *every* query, not just the most recent. A new, separate
3051 store pool is used for this. It can be recovered when all lookup caches are
3052 flushed. Lookups now release memory at the end of their result strings.
3053 This has involved some general refactoring of the lookup sources.
3055 7. Some code has been added to the store_xxx() functions to reduce the amount
3056 of flapping under certain conditions.
3058 8. log_incoming_interface used to affect only the <= reception log lines. Now
3059 it causes the local interface and port to be added to several more SMTP log
3060 lines, for example "SMTP connection from", and rejection lines.
3062 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3064 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3066 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3067 could be overwritten at the end of the current message (or the start of a
3068 new message if it was set in a HELO ACL). The value is now preserved for
3069 the duration of the SMTP connection.
3071 12. If a transport had a headers_rewrite setting, and a matching header line
3072 contained an unqualified address, that address was qualified, even if it
3073 did not match any rewriting rules. The underlying bug was that the values
3074 of the flags that permit the existence of unqualified sender and recipient
3075 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3076 non-local messages, and by -bnq for local messages) were not being
3077 preserved with the message after it was received.
3079 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3080 "next input=" as part of the text comprising the host identity instead of
3081 the correct text. The code was using the same buffer for two different
3082 strings. However, depending on which order the printing function evaluated
3083 its arguments, the bug did not always show up. Under Linux, for example, my
3084 test suite worked just fine.
3086 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3087 doesn't work with some older versions of Perl. It has been changed to "my",
3088 which in any case is probably the better facility to use.
3090 15. A really picky compiler found some instances of statements for creating
3091 error messages that either had too many or two few arguments for the format
3094 16. The size of the buffer for calls to the DNS resolver has been increased
3095 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3096 for addresses that have a lot of PTR records. This alleviates a problem; it
3097 does not fully solve it.
3099 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3100 buffer now truncates the list and logs the incident, which is the same
3101 action as happens when Exim is looking up a host name and its aliases.
3102 Previously in this situation something unpredictable would happen;
3103 sometimes it was "internal error: store_reset failed".
3105 18. If a server dropped the connection unexpectedly when an Exim client was
3106 using GnuTLS and trying to read a response, the client delivery process
3107 crashed while trying to generate an error log message.
3109 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3110 string, for example:
3112 warn message = H1: something\nH2: something
3114 the text was added as a single header line from Exim's point of view
3115 though it ended up OK in the delivered message. However, searching for the
3116 second and subsequent header lines using $h_h2: did not work. This has been
3117 fixed. Similarly, if a system filter added multiple headers in this way,
3118 the routers could not see them.
3120 20. Expanded the error message when iplsearch is called with an invalid key to
3121 suggest using net-iplsearch in a host list.
3123 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3124 ACLs are no longer actually imposed (and a message to that effect is
3127 22. If a "gecos" field in a passwd entry contained escaped characters, in
3128 particular, if it contained a \" sequence, Exim got it wrong when building
3129 a From: or a Sender: header from that name. A second bug also caused
3130 incorrect handling when an unquoted " was present following a character
3131 that needed quoting.
3133 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3134 was not being matched caselessly.
3136 24. Arranged for all hyphens in the exim.8 source to be escaped with
3139 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3140 properly. Recipient callout cache records were still being keyed to include
3141 the sender, even when use_sender was set false. This led to far more
3142 callouts that were necessary. The sender is no longer included in the key
3143 when use_sender is false.
3145 26. Added "control = submission" modifier to ACLs.
3147 27. Added the ${base62d: operator to decode base 62 numbers.
3149 28. dnsdb lookups can now access SRV records.
3151 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3152 the configuration file.
3154 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3155 (-v) mode. This makes the output for a verbose queue run more intelligible.
3157 31. Added a use_postmaster feature to recipient callouts.
3159 32. Added the $body_zerocount variable, containing the number of binary zero
3160 bytes in the message body.
3162 33. The time of last modification of the "new" subdirectory is now used as the
3163 "mailbox time last read" when there is a quota error for a maildir
3166 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3168 35. Added +ignore_unknown as a special item in host lists.
3170 36. Code for decoding IPv6 addresses in host lists is now included, even if
3171 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3172 address was recognized as an IP address, but was then not correctly decoded
3173 into binary, causing unexpected and incorrect effects when compared with
3180 1. Very minor rewording of debugging text in manualroute to say "list of
3181 hosts" instead of "hostlist".
3183 2. If verify=header_syntax was set, and a header line with an unqualified
3184 address (no domain) and a large number of spaces between the end of the
3185 name and the colon was received, the reception process suffered a buffer
3186 overflow, and (when I tested it) crashed. This was caused by some obsolete
3187 code that should have been removed. The fix is to remove it!
3189 3. When running in the test harness, delay a bit after writing a bounce
3190 message to get a bit more predictability in the log output.
3192 4. Added a call to search_tidyup() just before forking a reception process. In
3193 theory, someone could use a lookup in the expansion of smtp_accept_max_
3194 per_host which, without the tidyup, could leave open a database connection.
3196 5. Added the variables $recipient_data and $sender_data which get set from a
3197 lookup success in an ACL "recipients" or "senders" condition, or a router
3198 "senders" option, similar to $domain_data and $local_part_data.
3200 6. Moved the writing of debug_print from before to after the "senders" test
3203 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3204 problems for message scanning, either using a data ACL, or using
3205 local_scan() because the Received: header was not generated till after they
3206 were called (in order to set the time as the time of reception completion).
3207 I have revised the way this works. The header is now generated after the
3208 body is received, but before the ACL or local_scan() are called. After they
3209 are run, the timestamp in the header is updated.
3215 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3216 before starting a queue runner without re-exec. This happened only when
3217 deliver_drop_privilege was set or when the Exim user was set to root. The
3218 effect of the bug was that timeouts during subsequent deliveries caused
3219 crashes instead of being properly handled. The handler is now left at its
3220 default (and expected) setting.
3222 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3223 message, again when deliver_drop_privilege is set or Exim is run as root.
3224 The bug described in (1) was not present in this case, but the tidying up
3225 of the other signals was missing. I have made the two cases consistent.
3227 3. The ignore_target_hosts setting on a manualroute router was being ignored
3228 for hosts that were looked up using the /MX notation.
3230 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3233 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3234 operated on the sender address. After changing the $sender_address to <>
3235 for the sender address verify, Exim was re-instated it as the original
3236 (before rewriting) address, but remembering that it had rewritten it, so it
3237 wasn't rewriting it again. This bug also had the effect of breaking the
3238 sender address verification caching when the sender address was rewritten.
3240 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3241 This has been changed so that if the ip literal address matches
3242 ignore_target_hosts, the router declines.
3244 7. Added expansion conditions match_domain, match_address, and match_local_
3245 part (NOT match_host).
3247 8. The placeholder for the Received: header didn't have a length field set.
3249 9. Added code to Exim itself and to exim_lock to test for a specific race
3250 condition that could lead to file corruption when using MBX delivery. The
3251 issue is with the lockfile that is created in /tmp. If this file is removed
3252 after a process has opened it but before that process has acquired a lock,
3253 there is the potential for a second process to recreate the file and also
3254 acquire a lock. This could lead to two Exim processes writing to the file
3255 at the same time. The added code performs the same test as UW imapd; it
3256 checks after acquiring the lock that its file descriptor still refers to
3257 the same named file.
3259 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3260 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3263 11. Added the smtp_active_hostname option. If used, this will typically be made
3264 to depend on the incoming interface address. Because $interface_address is
3265 not set up until the daemon has forked a reception process, error responses
3266 that can happen earlier (such as "too many connections") no longer contain
3269 12. If an expansion in a condition on a "warn" statement fails because a lookup
3270 defers, the "warn" statement is abandoned, and the next ACL statement is
3271 processed. Previously this caused the whole ACL to be aborted.
3273 13. Added the iplsearch lookup type.
3275 14. Added ident_timeout as a log selector.
3277 15. Added tls_certificate_verified as a log selector.
3279 16. Added a global option tls_require_ciphers (compare the smtp transport
3280 option of the same name). This controls incoming TLS connections.
3282 17. I finally figured out how to make tls_require_ciphers do a similar thing
3283 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3284 before starting the TLS session.
3286 18. Tabs are now shown as \t in -bP output.
3288 19. If the log selector return_path_on_delivery was set, Exim crashed when
3289 bouncing a message because it had too many Received: header lines.
3291 20. If two routers both had headers_remove settings, and the first one included
3292 a superfluous trailing colon, the final name in the first list and the
3293 first name in the second list were incorrectly joined into one item (with a
3294 colon in the middle).
3300 1. Added -C and -D options to the exinext utility, mainly to make it easier
3301 to include in the automated testing, but these could be helpful when
3302 multiple configurations are in use.
3304 2. The exinext utility was not formatting the output nicely when there was
3305 an alternate port involved in the retry record key, nor when there was a
3306 message id as well (for retries that were specific to a specific message
3307 and a specific host). It was also confused by IPv6 addresses, because of
3308 the additional colons they contain. I have fixed the IPv4 problem, and
3309 patched it up to do a reasonable job for IPv6.
3311 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3312 delivery, the log line now contains "pipelined" if PIPELINING was used.
3314 4. An SMTP transport process used to panic and die if the bind() call to set
3315 an explicit outgoing interface failed. This has been changed; it is now
3316 treated in the same way as a connect() failure.
3318 5. A reference to $sender_host_name in the part of a conditional expansion
3319 that was being skipped was still causing a DNS lookup. This no longer
3322 6. The def: expansion condition was not recognizing references to header lines
3323 that used bh_ and bheader_.
3325 7. Added the _cache feature to named lists.
3327 8. The code for checking quota_filecount in the appendfile transport was
3328 allowing one more file than it should have been.
3330 9. For compatibility with Sendmail, the command line option
3338 and sets the incoming protocol and host name (for trusted callers). The
3339 host name and its colon can be omitted when only the protocol is to be set.
3340 Note the Exim already has two private options, -pd and -ps, that refer to
3341 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3342 "s", but I don't think that's a major issue.
3344 10. A number of refactoring changes to the code, none of which should affect
3347 (a) The number of logging options was getting close to filling up the
3348 32-bit word that was used as a bit map. I have split them into two classes:
3349 those that are passed in the argument to log_write(), and those that are
3350 only ever tested independently outside of that function. These are now in
3351 separate 32-bit words, so there is plenty of room for expansion again.
3352 There is no change in the user interface or the logging behaviour.
3354 (b) When building, for example, log lines, the code previously used a
3355 macro that called string_cat() twice, in order to add two strings. This is
3356 not really sufficiently general. Furthermore, there was one instance where
3357 it was actually wrong because one of the argument was used twice, and in
3358 one call a function was used. (As it happened, calling the function twice
3359 did not affect the overall behaviour.) The macro has been replaced by a
3360 function that can join an arbitrary number of extra strings onto a growing
3363 (c) The code for expansion conditions now uses a table and a binary chop
3364 instead of a serial search (which was left over from when there were very
3365 few conditions). Also, it now recognizes conditions like "pam" even when
3366 the relevant support is not compiled in: a suitably worded error message is
3367 given if an attempt is made to use such a condition.
3369 11. Added ${time_interval:xxxxx}.
3371 12. A bug was causing one of the ddress fields not to be passed back correctly
3372 from remote delivery subprocesses. The field in question was not being
3373 subsequently used, so this caused to problems in practice.
3375 13. Added new log selectors queue_time and deliver_time.
3377 14. Might have fixed a bug in maildirsizefile handling that threw up
3378 "unexpected character" debug warnings, and recalculated the data
3379 unnecessarily. In any case, I expanded the warning message to give more
3382 15. Added the message "Restricted characters in address" to the statements in
3383 the default ACL that block characters like @ and % in local parts.
3385 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3386 Three changes have been made:
3388 (a) There was a serious bug; a negative response to MAIL caused the whole
3389 recipient domain to be cached as invalid, thereby blocking all messages
3390 to all local parts at the same domain, from all senders. This bug has
3391 been fixed. The domain is no longer cached after a negative response to
3392 MAIL if the sender used is not empty.
3394 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3397 (c) A new callout option, "use_sender" has been added for people who want
3398 the modified behaviour.
3404 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3407 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3408 indeed breaks things for older releases.
3410 3. Added additional logging to the case where there is a problem reading data
3411 from a filter that is running in a subprocess using a pipe, in order to
3412 try to track down a specific problem.
3414 4. Testing facility fudge: when running in the test harness and attempting
3415 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3416 getting "No route to host". Convert this to a timeout.
3418 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3421 6. Some OS don't have socklen_t but use size_t instead. This affects the
3422 fifth argument of getsockopt() amongst other things. This is now
3423 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3424 can be set for individual OS. I have set it for SunOS5, OSF1, and
3425 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3426 some earlier ones do not.
3428 7. Change 4.30/15 was not doing the test caselessly.
3430 8. The standard form for an IPv6 address literal was being rejected by address
3431 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3432 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3433 this, as well as the form without the "IPv6" on the front (but only when
3434 address literals are enabled, of course).
3436 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3438 10. Exim crashed if a message with an empty sender address specified by -f
3439 encountered a router with an errors_to setting. This could be provoked only
3440 by a command such as
3444 where an empty string was supplied; "<>" did not hit this bug.
3446 11. Installed PCRE release 4.5.
3448 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3449 remained set. It is now erased.
3451 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3452 times from message ids (which are base 36 rather than the normal 62).
3454 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3455 were being counted as actual protocol errors, and logged if the log
3456 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3457 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3458 and DATA following a set of rejected RCPTs do not count as protocol errors.
3459 In other words, Exim assumes they were pipelined, though this may not
3460 actually be the case. Of course, in all cases the client gets an
3461 appropriate error code.
3463 15. If a lookup fails in an ACL condition, a message about the failure may
3464 be available; it is used if testing the ACL cannot continue, because most
3465 such messages specify what the cause of the deferral is. However, some
3466 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3467 that caused an old message to be retained and used if a later statement
3468 caused a defer, replacing the real cause of the deferral.
3470 16. If an IP address had so many PTR records that the DNS lookup buffer
3471 was not large enough to hold them, Exim could crash while trying to process
3472 the truncated data. It now detects and logs this case.
3474 17. Further to 4.21/58, another change has been made: if (and only if) the
3475 first line of a message (the first header line) ends with CRLF, a bare LF
3476 in a subsequent header line has a space inserted after it, so as not to
3477 terminate the header.
3479 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3480 unnecessarily, used atoi() instead of strtol(), and didn't check the
3481 termination when getting file sizes from file names by regex.
3483 19. Completely re-implemented the support for maildirsize files, in the light
3484 of a number of problems with the previous contributed implementation
3485 (4.30/29). In particular:
3487 . If the quota is zero, the maildirsize file is maintained, but no quota is
3490 . If the maildir directory does not exist, it is created before any attempt
3491 to write a maildirsize file.
3493 . The quota value in the file is just a cache; if the quota is changed in
3494 the transport, the new value overrides.
3496 . A regular expression is available for excluding directories from the
3499 20. The autoreply transport checks the characters in options that define the
3500 message's headers; it allows continued headers, but it was checking with
3501 isspace() after an embedded newline instead of explicitly looking for a
3504 21. If all the "regular" hosts to which an address was routed had passed their
3505 expiry times, and had not reached their retry times, the address was
3506 bounced, even if fallback hosts were defined. Now Exim should go on to try
3509 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3510 equivalent code in the SMTP transport. Some hosts send humungous responses
3511 to HELO/EHLO, more than 1024 it seems.
3513 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3514 gives compiler warnings in some environments. I've now done it "properly",
3517 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3518 (because of problems with the built-in one) was declared to return uschar *
3519 instead of char *, causing compiler failure.
3521 25. Fixed a file descriptor leak when processing alias/forward files.
3523 26. Fixed a minor format string issue in dbfn.c.
3525 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3527 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3528 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3531 29. When a sender address is verified, it is cached, to save repeating the test
3532 when there is more than one recipient in a message. However, when the
3533 verification involves a callout, it is possible for different callout
3534 options to be set for different recipients. It is too complicated to keep
3535 track of this in the cache, so now Exim always runs a verification when a
3536 callout is required, relying on the callout cache for the optimization.
3537 The overhead is duplication of the address routing, but this should not be
3540 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3541 to be verified with callout=postmaster, and the main callout worked but the
3542 postmaster check failed, the verification correctly failed. However, if a
3543 subsequent RCPT command asked for sender verification *without* the
3544 postmaster check, incorrect caching caused this verification also to fail,
3547 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3548 it was not caching the DNS options (qualify_single, search_parents) that
3549 were used when the lookup failed. A subsequent lookup with different
3550 options therefore always gave the same answer, though there were cases
3551 where it should not have. (Example: a "domains = !$mx_any" option on a
3552 dnslookup router: the "domains" option is always processed without any
3553 widening, but the router might have qualify_single set.) Now Exim uses the
3554 cached value only when the same options are set.
3556 32. Added John Jetmore's "exipick" utility to the distribution.
3558 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3559 than a timeout (e.g. a certificate is required, and is not provided), an
3560 Exim server now closes the connection immediately. Previously it waited for
3561 the client to close - but if the client is SSL, it seems that they each
3562 wait for each other, leading to a delay before one of them times out.
3564 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3565 maintained 0.8.x compatibility because I don't think many are using it, and
3566 it is clearly obsolete.
3568 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3571 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3572 client certificate was expired. A simple patch fixes this, though I don't
3573 understand the full logic of why the verify callback is called multiple
3576 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3577 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3578 which causes problems with some clients (such as the Certicom SSL Plus
3579 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3580 disables the coutermeasure allowing Eudora to connect."
3582 38. Exim was not checking that a write() to a log file succeeded. This could
3583 lead to Bad Things if a log got too big, in particular if it hit a file
3584 size limit. Exim now panics and dies if it cannot write to a log file, just
3585 as it does if it cannot open a log file.
3587 39. Modified OS/Makefile-Linux so that it now contains
3589 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3591 The two -D definitions ensure that Exim is compiled with large file
3592 support, which makes it possible to handle log files that are bigger than
3595 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3596 instance) a domain was checked against a named list that involved a lookup,
3597 causing $domain_data to be set, then another domain was checked against the
3598 same list, then the first domain was re-checked, the value of $domain_data
3599 after the final check could be wrong. In particular, if the second check
3600 failed, it could be set empty. This bug probably also applied to
3603 41. The strip_trailing_dot option was not being applied to the address given
3604 with the -f command-line option.
3606 42. The code for reading a message's header from the spool was incrementing
3607 $received_count, but never initializing it. This meant that the value was
3608 incorrect (doubled) while delivering a message in the same process in which
3609 it was received. In the most common configuration of Exim, this never
3610 happens - a fresh exec is done - but it can happen when
3611 deliver_drop_privilege is set.
3613 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3614 it now includes up to 150 characters of the unexpected data in the log
3617 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3618 and building data strings. The size of both of these buffers was 10 000
3619 bytes - far larger than anybody would *ever* want, thought I. Needless to
3620 say, somebody hit the limit. I have increased the maximum line length to
3621 20 000 and the maximum data length of concatenated lines to 100 000. I have
3622 also fixed two bugs, because there was no checking on these buffers. Tsk,
3623 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3624 buffer is too small.
3626 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3627 lsearch lookups. Now it does.
3629 46. When parsing a route_list item in a manualroute router, a fixed-length
3630 buffer was used for the list of hosts. I made this 1024 bytes long,
3631 thinking that nobody would ever have a list of hosts that long. Wrong.
3632 Somebody had a whole pile of complicated expansion conditions, and the
3633 string was silently truncated, leading to an expansion error. It turns out
3634 that it is easier to change to an unlimited length (owing to other changes
3635 that have happened since this code was originally written) than to build
3636 structure for giving a limitation error. The length of the item that
3637 expands into the list of hosts is now unlimited.
3639 47. The lsearch lookup could not handle data where the length of text line was
3640 more than 4095 characters. Such lines were truncated, leading to shortened
3641 data being returned. It should now handle lines of any length.
3643 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3644 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3647 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3648 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3649 no longer happens in comment lines. A list of replacements is now placed
3650 at the head of all of the source files, except those whose only change is
3651 to replace PERL_COMMAND in the very first #! line.
3653 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3654 messages on the queue, with a bottom-up merge sort, using code contributed
3655 by Michael Haardt. This should make operations like -bp somewhat faster on
3656 large queues. It won't affect queue runners, except when queue_run_in_order
3659 51. Installed eximstats 1.31 in the distribution.
3661 52. Added support for SRV lookups to the dnslookup router.
3663 53. If an ACL referred to $message_body or $message_body_end, the value was not
3664 reset for any messages that followed in the same SMTP session.
3666 54. The store-handling optimization for building very long strings was not
3667 differentiating between the different store pools. I don't think this
3668 actually made any difference in practice, but I've tidied it.
3670 55. While running the routers to verify a sender address, $sender_address
3671 was still set to the sender address. This is wrong, because when routing to
3672 send a bounce to the sender, it would be empty. Therefore, I have changed
3673 it so that, while verifying a sender address, $sender_address is set to <>.
3674 (There is no change to what happens when verifying a recipient address.)
3676 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3677 target A or AAAA records (if not already returned) without resetting the
3678 qualify_single or search_parents options of the DNS resolver. These are
3679 inappropriate in this case because the targets of MX and SRV records must
3680 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3681 target that, when qualified, matched something in the local domain. These
3682 two options are now turned off when doing these lookups.
3684 57. It seems that at least some releases of Reiserfs (which does not have the
3685 concept of a fixed number of inodes) returns zero and not -1 for the
3686 number of available inodes. This interacted badly with check_spool_inodes,
3687 which assumed that -1 was the "no such thing" setting. What I have done is
3688 to check that the total number of inodes is greater than zero before doing
3689 the test of how many are available.
3691 58. When a "warn" ACL statement has a log_message modifier, the message is
3692 remembered, and not repeated. This is to avoid a lot of repetition when a
3693 message has many recipients that cause the same warning to be written.
3694 Howewer, Exim was preserving the list of already written lines for an
3695 entire SMTP session, which doesn't seem right. The memory is now reset if a
3696 new message is started.
3698 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3699 debugging output unless log_rewrite was also set.
3701 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3702 of (int)(handle) when we know that handle contains (void *)(-1).
3704 61. The Exim daemon panic-logs an error return when it closes the incoming
3705 connection. However "connection reset by peer" seems to be common, and
3706 isn't really an error worthy of noting specially, so that particular error
3709 62. When Exim is trying to find all the local interfaces, it used to panic and
3710 die if the ioctl to get the interface flags failed. However, it seems that
3711 on at least one OS (Solaris 9) it is possible to have an interface that is
3712 included in the list of interfaces, but for which you get a failure error
3713 for this call. This happens when the interface is not "plumbed" into a
3714 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3715 failure of the "get flags" call assumes that the interface is down.
3717 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3718 makes life easier for people who are doing arithmetic on fields extracted
3719 from dates, where you often get leading zeros that should not be
3720 interpreted as octal.
3722 64. Added qualify_domain to the redirect router, to override the global
3725 65. If a pathologically long header line contained very many addresses (the
3726 report of this problem mentioned 10 000) and each of them was rewritten,
3727 Exim could use up a very large amount of memory. (It kept on making new
3728 copies of the header line as it rewrote, and never released the old ones.)
3729 At the expense of a bit more processing, the header rewriting function has
3730 been changed so that it no longer eats memory in this way.
3732 66. The generation of the Received: header has been moved from the time that a
3733 message starts to be received, to the time that it finishes. The timestamp
3734 in the Received: header should now be very close to that of the <= log
3735 line. There are two side-effects of this change:
3737 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3738 logged header lines no longer include the local Received: line, because
3739 it has not yet been created. The same applies to a copy of the message
3740 that is returned to a non-SMTP sender when a message is rejected.
3742 (b) When a filter file is tested using -bf, no additional Received: header
3743 is added to the test message. After some thought, I decided that this
3746 This change does not affect the value of $received_for. It is still set
3747 after address rewriting, but before local_scan() is called.
3749 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3751 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3752 gave an unhelpful panic error message, and a defer error. I have managed to
3753 change this behaviour so that it now rejects any supplied certificate,
3754 which seems right, as the list of acceptable certificates is empty.
3756 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3757 gave an unhelpful defer error. I have not managed to make this reject any
3758 supplied certificates, but the error message it gives is "no certificate
3759 supplied", which is not helpful.
3761 70. exigrep's output now also includes lines that are not associated with any
3762 message, but which match the given pattern. Implemented by a patch from
3763 Martin Sluka, which also tidied up the Perl a bit.
3765 71. Recipient callout verification, like sender verification, was using <> in
3766 the MAIL FROM command. This isn't really the right thing, since the actual
3767 sender may affect whether the remote host accepts the recipient or not. I
3768 have changed it to use the actual sender in the callout; this means that
3769 the cache record is now keyed on a recipient/sender pair, not just the
3770 recipient address. There doesn't seem to be a real danger of callout loops,
3771 since a callout by the remote host to check the sender would use <>.
3772 [SEE ABOVE: changed after hitting problems.]
3774 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3775 temporary errors. However, in the case of such a code being given after
3776 the end of a data transmission (i.e. after ".") Exim was failing to write
3777 a retry record for the message. (Yes, there was some broken host that was
3778 actually sending 8xx at this point.)
3780 73. An unknown lookup type in a host list could cause Exim to panic-die when
3781 the list was checked. (An example that provoked this was putting <; in the
3782 middle of a list instead of at the start.) If this happened during a DATA
3783 ACL check, a -D file could be left lying around. This kind of configuration
3784 error no longer causes Exim to die; instead it causes a defer errror. The
3785 incident is still logged to the main and panic logs.
3787 74. Buglet left over from Exim 3 conversion. The message "too many messages
3788 in one connection" was written to the rejectlog but not the mainlog, except
3789 when address rewriting (yes!) was being logged.
3791 75. Added write_rejectlog option.
3793 76. When a system filter was run not as root (that is, when system_filter_user
3794 was set), the values of the $n variables were not being returned to the
3795 main process; thus, they were not subsequently available in the $sn
3798 77. Added +return_path_on_delivery log selector.
3800 78. A connection timeout was being treated differently from recipients deferred
3801 when testing hosts_max_try with a message that was older than the host's
3802 retry timeout. (The host should not be counted, thus allowing all hosts to
3803 be tried at least once before bouncing.) This may have been the cause of an
3804 occasionally reported bug whereby a message would remain on the queue
3805 longer than the retry timeout, but would be bounced if a delivery was
3806 forced. I say "may" because I never totally pinned down the problem;
3807 setting up timeout/retry tests is difficult. See also the next item.
3809 79. The ultimate address timeout was not being applied to errors that involved
3810 a combination of host plus message (for example, a timeout on a MAIL
3811 command). When an address resolved to a number of possible hosts, and they
3812 were not all tried for each delivery (e.g. because of hosts_max_try), a
3813 message could remain on the queue longer than the retry timeout.
3815 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3818 81. Fixed an obscure SMTP outgoing bug which required at least the following
3819 conditions: (a) there was another message waiting for the same server;
3820 (b) the server returned 5xx to all RCPT commands in the first message so
3821 that the message was not completed; (c) the server dropped the connection
3822 or gave a negative response to the RSET that Exim sends to abort the
3823 transaction. The observed case was a dropped connection after DATA that had
3824 been sent in pipelining mode. That is, the server had advertised PIPELINING
3825 but was not implementing it correctly. The effect of the bug was incorrect
3826 behaviour, such as trying another host, and this could lead to a crash.
3832 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3833 and daemon.c were passed as pointers to ints; they should have been
3834 pointers to socklen_t variables (which are typically unsigned ints).
3836 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3839 3. Fixed a really odd bug that affected only the testing scheme; patching a
3840 certain fixed string in the binary changed the value of another string that
3841 happened to be identical to the end of the original first string.
3843 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3844 name", it returns that address as the IP address. On some operating
3845 systems (e.g. Solaris), it also passes back the IP address string as the
3846 "host name". However, on others (e.g. Linux), it passes back an empty
3847 string. Exim wasn't checking for this, and was changing the host name to an
3848 empty string, assuming it had been canonicized.
3850 5. Although rare, it is permitted to have more than one PTR record for a given
3851 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3852 all the names associated with an address, because they do in Solaris.
3853 However, it seems that they do not in Linux for data that comes from the
3854 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3855 I found this out when I moved to a new Linux workstation and tried to run
3856 the Exim test suite.
3858 To get round this problem I have changed the code so that it now does its
3859 own call to the DNS to look up PTR records when searching for a host name.
3860 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3861 addresses that are only in /etc/hosts are still found.
3863 This behaviour is, however, controlled by an option called host_lookup_
3864 order, which defaults to "bydns:byaddr". If people want to use the other
3865 order, or indeed, just use one or the other means of lookup, they can
3866 specify it in this variable.
3868 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3869 some operating systems, this comes back from gethostbyaddr() as an empty
3870 string, and this is what Exim used to test for. However, it seems that in
3871 other systems, "." is yielded. Exim now tests for this case too.
3873 7. The values of check_spool_space and check_log_space are now held internally
3874 as a number of kilobytes instead of an absolute number of bytes. If a
3875 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3876 kilobyte. This means that much larger values can be stored.
3878 8. Exim monitor: an attempt to get the action menu when not actually pointing
3879 at a message produces an empty menu entitled "No message selected". This
3880 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3881 no entries in it ("Shell widget menu has zero width and/or height"). So I
3882 have added a single, blank menu entry in this case.
3884 9. Added ${quote_local_part.
3886 10. MIME decoding is now applied to the contents of Subject: header lines when
3889 11. Now that a reference to $sender_host_address automatically causes a reverse
3890 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3891 host lookup before query-style lookups in lists that might use this
3892 variable. This has therefore been abolished, and the "net-" prefix is no
3893 longer necessary for query-style lookups.
3895 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3896 have been a typo for LFLAGS, so it has been changed.
3898 13. The install script calls Exim with "-C /dev/null" in order to find the
3899 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3900 to be output. Howeve, since Exim outputs its version number before the
3901 error, it didn't break the script. It just looked ugly. I fixed this by
3902 always allowing "-C /dev/null" if the caller is root.
3904 14. Ignore overlarge ACL variable number when reading spool file - insurance
3905 against a later release with more variables having written the file.
3907 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3908 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3909 this, as well as the form without the "IPv6" on the front.
3911 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3912 OS/Makefile-Darwin file.
3914 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3915 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3917 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3918 variable. However, some OS use ints for inodes. I've added cast to long int
3919 to get rid of the compiler warning.
3921 19. I had forgotten to lock out "/../" in configuration file names when
3922 ALT_CONFIG_PREFIX was set.
3924 20. Routers used for verification do not need to specify transports. However,
3925 if such a router generated a host list, and callout was configured, Exim
3926 crashed, because it could not find a port number from the (non-existent)
3927 transport. It now assumes port 25 in this circumstance.
3929 21. Added the -t option to exigrep.
3931 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3932 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3933 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3934 them is set, LOOKUP_LSEARCH is forced.
3936 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3938 24. Added sender and host information to the "rejected by local_scan()" log
3939 line; previously there was no indication of these.
3941 25. Added .include_if_exists.
3943 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3944 Linux. It turns out that not all file systems support this. Apparently some
3945 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3946 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3947 not supported on the file descriptor, is now ignored when Exim is trying to
3948 sync a directory. This applies only to Linux.
3950 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3952 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3953 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3956 29. Added support for maildirsize files from supplied patch (modified a bit).
3958 30. The use of :fail: followed by an empty string could lead Exim to respond to
3959 sender verification failures with (e.g.):
3961 550 Verification failed for <xxx>
3962 550 Sender verify failed
3964 where the first response line was missing the '-' that indicates it is not
3965 the final line of the response.
3967 31. The loop for finding the name of the user that called Exim had a hardwired
3968 limit of 10; it now uses the value of finduser_retries, which is used for
3969 all other user lookups.
3971 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3974 33. Exim was neglecting to zero errno before one call of strtol() when
3975 expanding a string and expecting an integer value. On some systems this
3976 resulted in spurious "integer overflow" errors. Also, it was casting the
3977 result into an int without checking.
3979 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3980 did not work. The code looks as if it has *never* worked, though it appears
3981 to have been documented since at least releast 1.62. I have made it work.
3983 35. The "timeout_DNS" error in retry rules, also documented since at least
3984 1.62, also never worked. As it isn't clear exactly what this means, and
3985 clearly it isn't a major issue, I have abolished the feature by treating it
3986 as "timeout", and writing a warning to the main and panic logs.
3988 36. The display of retry rules for -brt wasn't always showing the error code
3991 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3992 timeout_connect_A, timeout_connect_MX.
3994 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3995 to the empty sender.
3997 39. The daemon was not analysing the content of -oX till after it had closed
3998 stderr and disconnected from the controlling terminal. This meant that any
3999 syntax errors were only noted on the panic log, and the return code from
4000 the command was 0. By re-arranging the code a little, I've made the
4001 decoding happen first, so such errors now appear on stderr, and the return
4002 code is 1. However, the actual setting up of the sockets still happens in
4003 the disconnected process, so errors there are still only recorded on the
4006 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4007 connections (as happens on some IP stacks) was logged at start up time as
4008 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4009 it from "IPv6 and IPv4", which means that two separate sockets are being
4012 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4013 says whether AF_INET or AF_INET6 was passed as an argument.
4015 42. Exiwhat output was messed up when time zones were included in log
4018 43. Exiwhat now gives more information about the daemon's listening ports,
4019 and whether -tls-on-connect was used.
4021 44. The "port" option of the smtp transport is now expanded.
4023 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4024 silently ignored. Now an error message is written to the main and panic
4027 46. There's a new ACL modifier called "logwrite" which writes to a log file
4028 as soon as it is encountered.
4030 47. Added $local_user_uid and $local_user_gid at routing time.
4032 48. Exim crashed when trying to verify a sender address that was being
4035 49. Exim was recognizing only a space character after ".include". It now also
4036 recognizes a tab character.
4038 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4039 extracting the relevant information from the specification. The man page no
4040 longer contains scrambled data for the -d option, and I've added a section
4041 at the front about calling Exim under different names.
4043 51. Added "extra_headers" argument to the "mail" command in filter files.
4045 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4048 53. Installed eximstats 1.29.
4050 54. Added transport_filter_timeout as a generic transport option.
4052 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4053 Cc: header lines. This was required by RFC 822, but it not required by RFC
4056 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4057 incoming messages that did not have them. Now it does so only if the
4058 message originates locally, that is, if there is no associated remote host
4059 address. When Resent- header lines are present, this applies to the Resent-
4060 lines rather than the non-Resent- lines.
4062 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4063 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4065 58. Messages for configuration errors now include the name of the main
4066 configuration file - useful now that there may be more than one file in a
4067 list (.included file names were always shown).
4069 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4070 for those rare installations that do not start the daemon as root or run it
4071 setuid root. I've cut out the call to initgroups() if the daemon is not
4074 60. The Exim user and group can now be bound into the binary as text strings
4075 that are looked up at the start of Exim's processing.
4077 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4079 62. Added $mailstore_basename variable.
4081 63. Installed patch to sieve.c from Michael Haardt.
4083 64. When Exim failed to open the panic log after failing to open the main log,
4084 the original message it was trying to log was written to stderr and debug
4085 output, but if they were not available (the usual case in production), it
4086 was lost. Now it is written to syslog before the two lines that record the
4087 failures to open the logs.
4089 65. Users' Exim filters run in subprocesses under the user's uid. It is
4090 possible for a "deliver" command or an alias in a "personal" command to
4091 provoke an address rewrite. If logging of address rewriting is configured,
4092 this fails because the process is not running as root or exim. There may be
4093 a better way of dealing with this, but for the moment (because 4.30 needs
4094 to be released), I have disabled address rewrite logging when running a
4095 filter in a non-root, non-exim process.
4101 1. The buildconfig auxiliary program wasn't quoting the value set for
4102 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4103 not defined. This bug was masked in 4.22 by the effect that was fixed in
4106 2. Some messages that were rejected after a message id was allocated were
4107 shown as "incomplete" by exigrep. It no longer does this for messages that
4108 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4110 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4111 have allow_domain_literals set, the ID did not get logged in the <= line.
4112 Domain literals are now always recognized in Message-ID: header lines.
4114 4. The first argument for a ${extract expansion item is the key name or field
4115 number. Leading and trailing spaces in this item were not being ignored,
4116 causing some misleading effects.
4118 5. When deliver_drop_privilege was set, single queue runner processes started
4119 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4120 same command in the process it spins off) were not dropping privilege.
4122 6. When the daemon running as "exim" started a queue runner, it always
4123 re-executed Exim in the spun-off process. This is a waste of effort when
4124 deliver_drop_privilege is set. The new process now just calls the
4125 queue-runner function directly.
4131 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4134 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4135 ignored. Though the use of -oP was forcing the writing of a pid file, it
4136 was always written to the default place.
4138 3. If the message "no IP address found for host xxxx" is generated during
4139 incoming verification, it is now followed by identification of the incoming
4140 connection (so you can more easily find what provoked it).
4142 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4144 5. Added some features to "harden" Exim a bit more against certain attacks:
4146 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4147 be put in Local/Makefile. This is like the never_users runtime option,
4148 but it cannot be overridden. The default setting is "root".
4150 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4151 prefix string with which any file named in a -C command line option
4154 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4155 is retained for -C and -D only if the caller of Exim is root. Without
4156 it, the exim user may also use -C and -D and retain privilege.
4158 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4159 command line option is disabled.
4161 6. Macro names set by the -D option must start with an upper case letter, just
4162 like macro names defined in the configuration file.
4164 7. Added "dereference=" facility to LDAP.
4166 8. Two instances of the typo "uknown" in the source files are fixed.
4168 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4169 the Configure-Makefile script screwed up while processing it.
4171 10. Incorporated PCRE 4.4.
4173 11. The SMTP synchronization check was not operating right at the start of an
4174 SMTP session. For example, it could not catch a HELO sent before the client
4175 waited for the greeting. There is now a check for outstanding input at the
4176 point when the greeting is written. Because of the duplex, asynchronous
4177 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4178 way, but not yet received, when the check is performed.
4180 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4181 on TCP/IP sockets, because this apparently causes some broken clients to
4184 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4185 unchanged) from the Cygwin maintainer.
4187 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4188 was supported instead of testing for mbx. Effectively a typo.
4190 15. The spa authenticator server code was not checking that the input it
4191 received was valid base64.
4193 16. The debug output line for the "set" modifier in ACLs was not showing the
4194 name of the variable that was being set.
4196 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4198 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4199 lookup on its own. Something else was needed to trigger the lookup. For
4200 example, a match in host_lookup or the need for a host name in a host list.
4201 Now, if $sender_host_name is referenced and the host name has not yet been
4202 looked up, a lookup is performed. If the lookup fails, the variable remains
4203 empty, and $host_lookup_failed is set to "1".
4205 19. Added "eqi" as a case-independent comparison operator.
4207 20. The saslauthd authentication condition could segfault if neither service
4208 nor realm was specified.
4210 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4211 error message that was logged was misleading, and incoming SMTP
4212 connections were dropped. The message is now more accurate, and temporary
4213 errors are given to SMTP connections.
4215 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4216 (except RSET) with a 421 error, until QUIT is received. However, it was
4217 failing to send a response to QUIT.
4219 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4220 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4221 ACL is now run after the helo_try_verify_hosts code.
4223 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4224 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4225 case-independent, so other case variants are also recognized). Apparently
4226 some systems use these upper case variants.
4228 25. If more than two messages were waiting for the same host, and a transport
4229 filter was specified for the transport, Exim sent two messages over the
4230 same TCP/IP connection, and then failed with "socket operation on non-
4231 socket" when it tried to send the third.
4233 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4235 27. The extern definition of crypt16() in expand.c was not being excluded when
4236 the OS had its own crypt16() function.
4238 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4239 as a preferred synonym for return_size_limit, both as an option and as an
4242 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4244 30. Changed the default configuration ACL to relax the local part checking rule
4245 for addresses that are not in any local domains. For these addresses,
4246 slashes and pipe symbols are allowed within local parts, but the sequence
4247 /../ is explicitly forbidden.
4249 31. SPA server authentication was not clearing the challenge buffer before
4252 32. log_message in a "warn" ACL statement was writing to the reject log as
4253 well as to the main log, which contradicts the documentation and doesn't
4254 seem right (because no rejection is happening). So I have stopped it.
4256 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4257 However, I am unable to do any testing of this.
4259 34. Fixed an infelicity in the appendfile transport. When checking directories
4260 for a mailbox, to see if any needed to be created, it was accidentally
4261 using path names with one or more superfluous leading slashes; tracing
4262 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4264 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4265 contents are added to the log line that is written for every discarded
4266 recipient. (Previously a log_message setting was ignored.)
4268 36. The ${quote: operator now quotes the string if it is empty.
4270 37. The install script runs exim in order to find its version number. If for
4271 some reason other than non-existence or emptiness, which it checks, it
4272 could not run './exim', it was installing it with an empty version number,
4273 i.e. as "exim-". This error state is now caught, and the installation is
4276 38. An argument was missing from the function that creates an error message
4277 when Exim fails to connect to the socket for saslauthd authentication.
4278 This could cause Exim to crash, or give a corrupted message.
4280 39. Added isip, isip4, and isip6 to ${if conditions.
4282 40. The ACL variables $acl_xx are now saved with the message, and can be
4283 accessed later in routers, transports, and filters.
4285 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4286 strings in the file are not string-expanded.
4288 42. If a MAIL command specified a SIZE value that was too large to fit into an
4289 int variable, the check against message_size_limit failed. Such values are
4290 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4291 day this will have to be increased, but I don't think I want to be around
4292 when emails are that large.
4299 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4300 iconv() is not standard in FreeBSD.
4302 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4303 IPv6 enabled. The observed symptom was a segmentation fault on return from
4304 the function os_common_find_running_interfaces() in src/os.c.
4306 3. In the check_special_case() function in daemon.c I had used "errno" as an
4307 argument name, which causes warnings on some systems. This was basically a
4308 typo, since it was named "eno" in the comments!
4310 4. The code that waits for the clock to tick (at a resolution of some fraction
4311 of a second) so as to ensure message-id uniqueness was always waiting for
4312 at least one whole tick, when it could have waited for less. [This is
4313 almost certainly not relevant at current processor speeds, where it is
4314 unlikely to ever wait at all. But we try to future-proof.]
4316 5. The function that sleeps for a time interval that includes fractions of a
4317 second contained a race. It did not block SIGALRM between setting the
4318 timer, and suspending (a couple of lines later). If the interval was short
4319 and the sigsuspend() was delayed until after it had expired, the suspension
4320 never ended. On busy systems this could lead to processes getting stuck for
4323 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4324 process, before it forks any delivery processes. The open lookup caching
4325 mechanism meant that the open file or database connection was passed into
4326 the delivery process. The problem was that delivery processes always tidy
4327 up cached lookup data. This could cause a problem for the next delivery
4328 process started by the queue runner, because the external queue runner
4329 process does not know about the closure. So the next delivery process
4330 still has data in the lookup cache. In the case of a file lookup, there was
4331 no problem because closing a file descriptor in a subprocess doesn't affect
4332 the parent. However, if the lookup was caching a connection to a database,
4333 the connection was closed, and the second delivery process was likely to
4334 see errors such as "PGSQL: query failed: server closed the connection
4335 unexpectedly". The problem has been fixed by closing all cached lookups
4336 in a queue runner before running a delivery process.
4338 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4339 seem to have the "const" qualifier which it has on other OS. I've
4342 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4343 *of the same type* (client or server) with the same public name that an
4344 error should be diagnosed.
4346 9. When Exim looked up a host name for an IP address, but failed to find the
4347 original IP address when looking up the host name (a safety check), it
4348 output the message "<ip address> does not match any IP for NULL", which was
4349 confusing, to say the least. The bug was that the host name should have
4350 appeared instead of "NULL".
4352 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4353 user that is built into the binary, and the -C or -D options is used, root
4354 privilege is dropped before the configuration file is read. In addition,
4355 logging is switched to stderr instead of the normal log files. If the
4356 configuration then re-defines the Exim user, the unprivileged environment
4357 is probably not what is expected, so Exim logs a panic warning message (but
4360 However, if deliver_drop_privilege is set, the unprivileged state may well
4361 be exactly what is intended, so the warning has been cut out in that case,
4362 and Exim is allowed to try to write to its normal log files.
4368 1. smtp_return_error_details was not giving details for temporary sender
4369 or receiver verification errors.
4371 2. Diagnose a configuration error if two authenticators have the same public
4374 3. Exim used not to create the message log file for a message until the first
4375 delivery attempt. This could be confusing when incoming messages were held
4376 for policy or load reasons. The message log file is now created at the time
4377 the message is received, and an initial "Received" line is written to it.
4379 4. The automatically generated man page for command line options had a minor
4380 bug that caused no ill effects; however, a more serious problem was that
4381 the procedure for building the man page automatically didn't always
4382 operate. Consequently, release 4.20 contains an out-of-date version. This
4383 shouldn't happen again.
4385 5. When building Exim with embedded Perl support, the script that builds the
4386 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4387 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4389 6. The freeze_tell option was not being used for messages that were frozen on
4390 arrival, either by an ACL or by local_scan().
4392 7. Added the smtp_incomplete_transaction log selector.
4394 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4395 was accepting AUTH without a new EHLO.
4397 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4398 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4401 10. Logging of TCP/IP connections (when configured) now happens in the main
4402 daemon process instead of the child process, so that the TCP/IP connection
4403 count is more accurate (but it can never be perfect).
4405 11. The use of "drop" in a nested ACL was not being handled correctly in the
4406 outer ACL. Now, if condition failure induced by the nested "drop" causes
4407 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4408 or "require"), the connection is dropped.
4410 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4411 that yield "discard" can now be used with an "accept" or a "discard" verb,
4412 but an error is generated for any others (because I can't see a useful way
4413 to define what should happen).
4415 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4416 are now processed in the same way as lines in the Exim configuration file.
4417 In particular, continuation lines are supported.
4419 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4421 15. Added -ti meaning -t -i.
4423 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4424 domains, and warn by logging if others are found.
4426 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4427 returned from GIFCONF when Exim is trying to find the list of interfaces on
4428 a host. The code in os.c has been modified to copy each ifreq to an aligned
4429 structure in all cases.
4431 Also, in some cases, the returned ifreq's were being copied to a 'struct
4432 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4433 means the last couple of bytes of an IPv6 address could be chopped if the
4434 ifreq contained only a normal sockaddr (14 bytes storage).
4436 18. Named domain lists were not supported in the hosts_treat_as_local option.
4437 An entry such as +xxxx was not recognized, and was treated as a literal
4440 19. Ensure that header lines added by a DATA ACL are included in the reject log
4441 if the ACL subsequently rejects the message.
4443 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4444 MD5 (which is deprecated).
4446 21. When testing a filter file using -bf, Exim was writing a message when it
4447 took the sender from a "From " line in the message, but it was not doing so
4448 when it took $return_path from a Return-Path: header line. It now does.
4450 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4451 with a valid header line field name (a series of printing characters
4452 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4454 23. Changed "disc" in the source to "disk" to conform to the documentation and
4455 the book and for uniformity.
4457 24. Ignore Sendmail's -Ooption=value command line item.
4459 25. When execve() failed while trying to run a command in a pipe transport,
4460 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4461 could be confused with a return value of 69 from the command itself. This
4462 has been changed to 127, the value the shell returns if it is asked to run
4463 a non-existent command. The wording for the related log line suggests a
4464 non-existent command as the problem.
4466 26. If received_header_text expands to an empty string, do not add a Received:
4467 header line to the message. (Well, it adds a token one on the spool, but
4468 marks it "old" so that it doesn't get used or transmitted.)
4470 27. Installed eximstats 1.28 (addition of -nt option).
4472 28. There was no check for failure on the call to getsockname() in the daemon
4473 code. This can fail if there is a shortage of resources on the system, with
4474 ENOMEM, for example. A temporary error is now given on failure.
4476 29. Contrary to the C standard, it seems that in some environments, the
4477 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4478 program. Exim now does this explicitly; it affects the formatting of
4479 timestamps using strftime().
4481 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4482 complaints. I've now initialized all the variables, to avoid this.
4484 32. Header lines added by a system filter were not being "seen" during
4485 transport-time rewrites.
4487 33. The info_callback() function passed to OpenSSL is set up with type void
4488 (*)(SSL *, int, int), as described somewhere. However, when calling the
4489 function (actually a macro) that sets it up, the type void(*)() is
4490 expected. I've put in a cast to prevent warnings from picky compilers.
4492 34. If a DNS black list lookup found a CNAME record, but there were no A
4493 records associated with the domain it pointed at, Exim crashed.
4495 35. If a DNS black list lookup returned more than one A record, Exim ignored
4496 all but the first. It now scans all returned addresses if a particular IP
4497 value is being sought. In this situation, the contents of the
4498 $dnslist_value variable are a list of all the addresses, separated by a
4501 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4502 to accept a host name and all its aliases if the forward lookup for any of
4503 them yielded the IP address of the incoming connection. Now it accepts only
4504 those names whose forward lookup yields the correct IP address. Any other
4505 names are discarded. This closes a loophole whereby a rogue DNS
4506 administrator could create reverse DNS records to break through a
4507 wildcarded host restriction in an ACL.
4509 37. If a user filter or a system filter that ran in a subprocess used any of
4510 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4511 the wrong values were passed to the pipe command ($thisaddress had the
4512 value of $0, $0 had the value of $1, etc). This bug was introduced by
4513 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4515 38. Improved the line breaking for long SMTP error messages from ACLs.
4516 Previously, if there was no break point between 40 and 75 characters, Exim
4517 left the rest of the message alone. Two changes have been made: (a) I've
4518 reduced the minimum length to 35 characters; (b) if it can't find a break
4519 point between 35 and 75 characters, it looks ahead and uses the first one
4520 that it finds. This may give the occasional overlong line, but at least the
4521 remaining text gets split now.
4523 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4524 file descriptors might be low, and that setting 1000 would always raise it.
4525 It turns out that in some environments, the limit is already over 1000 and
4526 that lowering it causes trouble. So now Exim takes care not to decrease it.
4528 40. When delivering a message, the value of $return_path is set to $sender_
4529 address at the start of routing (routers may change the value). By an
4530 oversight, this default was not being set up when an address was tested by
4531 -bt or -bv, which affected the outcome if any router or filter referred to
4534 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4535 log only when "message" or "log_message" are set. However, if one of the
4536 conditions was an address verification, or a call to a nested ACL, the
4537 messages generated by the underlying test were being passed through. This
4538 no longer happens. The underlying message is available in $acl_verify_
4539 message for both "message" and "log_message" expansions, so it can be
4540 passed through if needed.
4542 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4543 new expansion $bh_ to give the encoded byte string without charset
4544 translation. Translation happens only if iconv() is available; HAVE_ICONV
4545 indicates this at build time. HEADERS_CHARSET gives the charset to
4546 translate to; headers_charset can change it in the configuration, and
4547 "headers charset" can change it in an individual filter file.
4549 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4550 that creates RFC 2047 encoded "words" labels them as that charset instead
4551 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4552 expansion operator; (ii) when Exim creates a From: line for a local
4553 message; (iii) when a header line is rewritten to include a "phrase" part.
4555 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4556 buggy, causing it to skip the first lines of messages whose message ID
4557 ended in 'D'. This would not have bitten before Exim release 4.14, because
4558 message IDs were unlikely to end in 'D' before then. The effect was to have
4559 incorrect size information for certain domains.
4561 45. #include "config.h" was missing at the start of the crypt16.c module. This
4562 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4565 46. If there was a timeout during a "random" callout check, Exim treated it as
4566 a failure of the random address, and carried on sending RSET and the real
4567 address. If the delay was just some slowness somewhere, the response to the
4568 original RCPT would be taken as a response to RSET and so on, causing
4569 mayhem of various kinds.
4571 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4572 when I implemented it. It didn't allow for the fact that some option values
4573 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4574 the right test for positive values.
4576 48. Domain names in DNS records are case-independent. Exim always looks them up
4577 in lower case. Some resolvers return domain names in exactly the case they
4578 appear in the zone file, that is, they may contain uppercase letters. Not
4579 all resolvers do this - some return always lower case. Exim was treating a
4580 change of case by a resolver as a change of domain, similar to a widening
4581 of a domain abbreviation. This triggered its re-routing code and so it was
4582 trying to route what was effectively the same domain again. This normally
4583 caused routing to fail (because the router wouldn't handle the domain
4584 twice). Now Exim checks for this case specially, and just changes the
4585 casing of the domain that it ultimately uses when it transmits the message
4588 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4591 50. If a filter generated a file delivery with a non-absolute name (possible if
4592 no home directory exists for the router), the forbid_file option was not
4595 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4596 the existing equality matching.
4598 52. Exim was using ints instead of ino_t variables in some places where it was
4599 dealing with inode numbers.
4601 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4602 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4603 called TMPDIR, and if it finds it is different, it changes its value.
4605 54. The smtp_printf() function is now made available to local_scan() so
4606 additional output lines can be written before returning. There is also an
4607 smtp_fflush() function to enable the detection of a dropped connection.
4608 The variables smtp_input and smtp_batched_input are exported to
4611 55. Changed the default runtime configuration: the message "Unknown user"
4612 has been removed from the ACL, and instead placed on the localuser router,
4613 using the cannot_route_message feature. This means that any verification
4614 failures that generate their own messages won't get overridden. Similarly,
4615 the "Unrouteable address" message that was in the ACL for unverifiable
4616 relay addresses has also been removed.
4618 56. Added hosts_avoid_esmtp to the smtp transport.
4620 57. The exicyclog script was not checking for the esoteric option
4621 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4622 will work only if exicyclog is run under the appropriate euid.
4624 58. Following a discussion on the list, the rules by which Exim recognises line
4625 endings on incoming messages have been changed. The -dropcr and drop_cr
4626 options are now no-ops, retained only for backwards compatibility. The
4627 following line terminators are recognized: LF CRLF CR. However, special
4628 processing applies to CR:
4630 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4631 nor a local message in the state where . is a terminator.
4633 (ii) If a bare CR is encountered in a header line, an extra space is added
4634 after the line terminator so as not to end the header. The reasoning
4635 behind this is that bare CRs in header lines are most likely either
4636 to be mistakes, or people trying to play silly games.
4638 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4639 was being incorrectly given as 18 bytes larger than it should have been.
4640 This is a VOB (very old bug).
4642 60. This may never have affected anything current, but just in case it has:
4643 When the local host is found other than at the start of a list of hosts,
4644 the local host, those with the same MX, and any that follow, are discarded.
4645 When the list in question was part of a longer list of hosts, the following
4646 hosts (not currently being processed) were also being discarded. This no
4647 longer happens. I'm not sure if this situation could ever has previously
4650 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4653 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4654 header. This is something that is recommended in a new Internet Draft, and
4655 is something that is documented as being done by Sendmail. There are two
4656 possible values. For messages generated by the autoreply transport, Exim
4659 Auto-Submitted: auto-replied
4661 whereas for all other generated messages (e.g. bounces) it adds
4663 Auto-Submitted: auto-generated
4665 63. The "personal" condition in filters now includes a test for the
4666 Auto-Submitted: header. If it contains the string "auto-" the message it
4667 not considered personal.
4669 64. Added rcpt_include_affixes as a generic transport option.
4671 65. Added queue_only_override (default true).
4673 66. Added the syslog_duplication option.
4675 67. If what should have been the first header line of a message consisted of
4676 a space followed by a colon, Exim was mis-interpreting it as a header line.
4677 It isn't of course - it is syntactically invalid and should therefore be
4678 treated as the start of the message body. The misbehaviour could have
4679 caused a number of strange effects, including loss of data in subsequent
4680 header lines, and spool format errors.
4682 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4683 client host had authenticated. This control can now be exercised by an ACL
4684 for more flexibility.
4686 69. By default, callouts do not happen when testing with -bh. There is now a
4687 variant, -bhc, which does actually run the callout code, including
4688 consulting and updating the callout cache.
4690 70. Added support for saslauthd authentication, courtesy of Alexander
4693 71. If statvfs() failed on the spool or log directories while checking their
4694 size for availability, Exim confusingly gave the error "space shortage".
4695 Furthermore, in debugging mode it crashed with a floating point exception.
4696 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4697 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4698 serious problem, Exim now writes to the main and panic logs when this
4699 happens, with details of the failure. It then refuses to accept the
4700 incoming message, giving the message "spool directory problem" or "log
4701 directory problem" with a 421 code for SMTP messages.
4703 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4704 0, 1, and 2 exist, because some OS complain for execs without them (see
4705 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4706 descriptors. However, the code omitted to check that the open succeeded,
4707 causing mysterious errors if for some reason the permissions on /dev/null
4708 got screwed. Now Exim writes a message to the main and panic logs, and
4709 bombs out if it can't open /dev/null.
4711 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4712 interact so that it is all more flexible. It is supposed to remain
4713 backwards compatible. Also added extra_local_interfaces.
4715 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4716 to bomb out with an assertion failure - to the client this appears as a
4717 connection drop. This problem occurs in the part of the code that was taken
4718 from the Samba project. Fortunately, the assertion is in a very simple
4719 function, so I have fixed this by reproducing the function inline in the
4720 one place where it is called, and arranging for authentication to fail
4721 instead of killing the process with assert().
4723 75. The SPA client code was not working when the server requested OEM rather
4724 than Unicode encoding.
4726 76. Added code to make require_files with a specific uid setting more usable in
4727 the case where statting the file as root fails - usually a non-root-mounted
4728 NFS file system. When this happens and the failure is EACCES, Exim now
4729 forks a subprocess and does the per-uid checking as the relevant uid.
4731 77. Added process_log_path.
4733 78. If log_file_path was not explicitly set, a setting of check_log_space or
4734 check_log_inodes was ignored.
4736 79. If a space check for the spool or log partitions fails, the incident is now
4737 logged. Of course, in the latter case the data may get lost...
4739 80. Added the %p formatting code to string_format() so that it can be used to
4740 print addresses in debug_print(). Adjusted all the address printing in the
4741 debugging in store.c to use %p rather than %d.
4743 81. There was a concern that a line of code in smtp_in.c could overflow a
4744 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4745 initially expressed, the concern was not well-founded, because trailing
4746 spaces are removed early. However, if the trailing spaces were followed by
4747 a NULL, they did not get removed, so the overflow was possible. Two fixes
4750 (a) I re-wrote the offending code in a cleaner fashion.
4751 (b) If an incoming SMTP command contains a NULL character, it is rejected
4754 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4755 runs initgroups(), so that if the Exim user is in any additional groups,
4756 they will be used during message reception.
4762 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob