1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.533 2007/09/28 13:04:16 tom Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
6 TK/01 Add preliminary DKIM support. Currently requires a forked version of
7 ALT-N's libdkim that I have put here:
8 http://duncanthrax.net/exim-experimental/
10 Note to Michael Haardt: I had to rename some vars in sieve.c. They
11 were called 'true' and it seems that C99 defines that as a reserved
12 keyword to be used with 'bool' variable types. That means you could
13 not include C99-style headers which use bools without triggering
14 build errors in sieve.c.
16 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
17 as mailq or other aliases. Changed the --help handling significantly
18 to do whats expected. exim_usage() emits usage/help information.
20 SC/01 Added the -bylocaldomain option to eximstats.
25 PH/01 Another patch from the Sieve maintainer.
27 PH/02 When an IPv6 address is converted to a string for single-key lookup
28 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
29 dots are used instead of colons so that keys in lsearch files need not
30 contain colons. This was done some time before quoting was made available
31 in lsearch files. However, iplsearch files do require colons in IPv6 keys
32 (notated using the quote facility) so as to distinguish them from IPv4
33 keys. This meant that lookups for IP addresses in host lists did not work
34 for iplsearch lookups.
36 This has been fixed by arranging for IPv6 addresses to be expressed with
37 colons if the lookup type is iplsearch. This is not incompatible, because
38 previously such lookups could never work.
40 The situation is now rather anomolous, since one *can* have colons in
41 ordinary lsearch keys. However, making the change in all cases is
42 incompatible and would probably break a number of configurations.
44 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
47 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
48 conversion specification without a maximum field width, thereby enabling
49 a rogue spamd server to cause a buffer overflow. While nobody in their
50 right mind would setup Exim to query an untrusted spamd server, an
51 attacker that gains access to a server running spamd could potentially
52 exploit this vulnerability to run arbitrary code as the Exim user.
54 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
55 $primary_hostname instead of what libspf2 thinks the hosts name is.
57 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
58 a directory entry by the name of the lookup key. Previously, if a
59 symlink pointed to a non-existing file or a file in a directory that
60 Exim lacked permissions to read, a lookup for a key matching that
61 symlink would fail. Now it is enough that a matching directory entry
62 exists, symlink or not. (Bugzilla 503.)
64 PH/03 The body_linecount and body_zerocount variables are now exported in the
67 PH/04 Added the $dnslist_matched variable.
69 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
70 This means they are set thereafter only if the connection becomes
73 PH/06 Added the client_condition to authenticators so that some can be skipped
74 by clients under certain conditions.
76 PH/07 The error message for a badly-placed control=no_multiline_responses left
77 "_responses" off the end of the name.
79 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
81 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
82 (without spaces) instead of just copying the configuration text.
84 PH/10 Added the /noupdate option to the ratelimit ACL condition.
86 PH/11 Added $max_received_linelength.
88 PH/12 Added +ignore_defer and +include_defer to host lists.
90 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
91 way in which PCRE > 7.0 is built.
93 PH/14 Implemented queue_only_load_latch.
95 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
96 MAIL command. The effect was to mangle the value on 64-bit systems.
98 PH/16 Another patch from the Sieve maintainer.
100 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
102 PH/18 If a system quota error occurred while trying to create the file for
103 a maildir delivery, the message "Mailbox is full" was not appended to the
104 bounce if the delivery eventually timed out. Change 4.67/27 below applied
105 only to a quota excession during the actual writing of the file.
107 PH/19 It seems that peer DN values may contain newlines (and other non-printing
108 characters?) which causes problems in log lines. The DN values are now
109 passed through string_printing() before being added to log lines.
111 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
112 and InterBase are left for another time.)
114 PH/21 Added message_body_newlines option.
116 PH/22 Guard against possible overflow in moan_check_errorcopy().
118 PH/23 POSIX allows open() to be a macro; guard against that.
120 PH/24 If the recipient of an error message contained an @ in the local part
121 (suitably quoted, of course), incorrect values were put in $domain and
122 $local_part during the evaluation of errors_copy.
128 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
129 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
132 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
133 issue a MAIL command.
135 PH/02 In an ACL statement such as
137 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
139 if a client was not listed at all, or was listed with a value other than
140 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
141 the condition was not true (as it should be), so access was not denied.
142 The bug was that the ! inversion was incorrectly passed on to the second
143 item. This has been fixed.
145 PH/03 Added additional dnslists conditions == and =& which are different from
146 = and & when the dns lookup returns more than one IP address.
148 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
149 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
151 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
152 FSYNC, which compiles an option called disable_fsync that allows for
153 bypassing fsync(). The documentation is heavily laced with warnings.
155 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
157 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
158 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
159 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
160 including adding "make clean"; (3) Added -fPIC when compiling the test
161 dynamically loaded module, to get rid of a warning.
163 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
164 message fails, move_frozen_messages = true and ignore_bounce_errors_after
165 = 0s. The bug is otherwise harmless.
167 PH/07 There was a bug in the dovecot authenticator such that the value of
168 $auth1 could be overwritten, and so not correctly preserved, after a
169 successful authentication. This usually meant that the value preserved by
170 the server_setid option was incorrect.
172 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
174 PH/09 Installed PCRE release 7.0.
176 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
177 run for batched SMTP input. It is now run at the start of every message
178 in the batch. While fixing this I discovered that the process information
179 (output by running exiwhat) was not always getting set for -bs and -bS
180 input. This is fixed, and it now also says "batched" for BSMTP.
182 PH/11 Added control=no_pipelining.
184 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
185 patch, slightly modified), and move the expansion of helo_data till after
186 the connection is made in the smtp transport (so it can use these
189 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
191 PH/14 Added log_selector = +pid.
193 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
195 PH/16 Add ${if forany and ${if forall.
197 PH/17 Added dsn_from option to vary the From: line in DSNs.
199 PH/18 Flush SMTP output before performing a callout, unless control =
200 no_callout_flush is set.
202 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
203 was true (the default) a successful delivery failed to delete the retry
204 item, thus causing premature timeout of the address. The bug is now
207 PH/20 Added hosts_avoid_pipelining to the smtp transport.
209 PH/21 Long custom messages for fakedefer and fakereject are now split up
210 into multiline reponses in the same way that messages for "deny" and
211 other ACL rejections are.
213 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
214 with slight modification.
216 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
217 draft, changing the syntax and factoring some duplicate code".
219 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
220 for deliveries of the second and subsequent messages over the same SMTP
223 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
224 ${reduce, with only minor "tidies".
226 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
228 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
229 expansion side effects.
231 PH/27 When a message times out after an over-quota error from an Exim-imposed
232 quota, the bounce message says "mailbox is full". This message was not
233 being given when it was a system quota that was exceeded. It now should
236 MH/03 Made $recipients available in local_scan(). local_scan() already has
237 better access to the recipient list through recipients_list[], but
238 $recipients can be useful in postmaster-provided expansion strings.
240 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
241 in the case of a MAIL command with additional options following the
242 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
243 were accidentally chopped off.
245 PH/29 SMTP synchronization checks are implemented when a command is read -
246 there is a check that no more input is waiting when there shouldn't be
247 any. However, for some commands, a delay in an ACL can mean that it is
248 some time before the response is written. In this time, more input might
249 arrive, invalidly. So now there are extra checks after an ACL has run for
250 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
251 pipelining has not been advertised.
253 PH/30 MH's patch to allow iscntrl() characters to be list separators.
255 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
256 returned in the SMTP response when smtp_return_error_details was false.
259 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
260 instead of the C I/O that was originally supplied, because problems were
263 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
264 Exim which did not show up earlier: it was assuming that a call to
265 SSL_CTX_set_info_callback() might give an error value. In fact, there is
266 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
267 was a macro that became an assignment, so it seemed to work. This has
268 changed to a proper function call with a void return, hence the compile
269 error. Exim's code has been fixed.
271 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
274 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
276 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
278 PH/37 If a message is not accepted after it has had an id assigned (e.g.
279 because it turns out to be too big or there is a timeout) there is no
280 "Completed" line in the log. When some messages of this type were
281 selected by exigrep, they were listed as "not completed". Others were
282 picked up by some special patterns. I have improved the selection
283 criteria to be more general.
285 PH/38 The host_find_failed option in the manualroute router can now be set
286 to "ignore", to completely ignore a host whose IP address cannot be
287 found. If all hosts are ignored, the behaviour is controlled by the new
288 host_all_ignored option.
290 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
291 homing or because of multiple MX records with /mx) generated more than
292 one IP address, and the following item turned out to be the local host,
293 all the secondary addresses of the first item were incorrectly removed
294 from the list, along with the local host and any following hosts (which
295 is what is supposed to happen).
297 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
298 whoever called Exim into the -H file. In the case of the daemon it was
299 behaving confusingly. When first started, it used values for whoever
300 started the daemon, but after a SIGHUP it used the Exim user (because it
301 calls itself on a restart). I have changed the code so that it now always
304 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
305 message are rejected with the same error (e.g. no authentication or bad
306 sender address), and a DATA command is nevertheless sent (as can happen
307 with PIPELINING or a stupid MUA), the error message that was given to the
308 RCPT commands is included in the rejection of the DATA command. This is
309 intended to be helpful for MUAs that show only the final error to their
312 PH/42 Another patch from the Sieve maintainer.
314 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
315 Eximstats - Fixed some broken HTML links and added missing column headers
317 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
318 columns for Rejects, Temp Rejects, Ham, and Spam rows.
320 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
322 PH/43 Yet another patch from the Sieve maintainer.
324 PH/44 I found a way to check for a TCP/IP connection going away before sending
325 the response to the final '.' that terminates a message, but only in the
326 case where the client has not sent further data following the '.'
327 (unfortunately, this is allowed). However, in many cases there won't be
328 any further data because there won't be any more messages to send. A call
329 to select() can be used: if it shows that the input is "ready", there is
330 either input waiting, or the socket has been closed. An attempt to read
331 the next input character can distinguish the two cases. Previously, Exim
332 would have sent an OK response which the client would never have see.
333 This could lead to message repetition. This fix should cure that, at
334 least in a lot of common cases.
336 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
337 advertised in response to EHLO.
343 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
344 fixed by 4.65/MH/01 (is this a record?) are fixed:
346 (i) An empty string was always treated as zero by the numeric comparison
347 operators. This behaviour has been restored.
349 (ii) It is documented that the numeric comparison operators always treat
350 their arguments as decimal numbers. This was broken in that numbers
351 starting with 0 were being interpreted as octal.
353 While fixing these problems I realized that there was another issue that
354 hadn't been noticed. Values of message_size_limit (both the global option
355 and the transport option) were treated as octal if they started with 0.
356 The documentation was vague. These values are now always treated as
357 decimal, and I will make that clear in the documentation.
363 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
364 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
367 MH/01 Don't check that the operands of numeric comparison operators are
368 integers when their expansion is in "skipping" mode (fixes bug
369 introduced by 4.64-PH/07).
371 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
372 child addresses, Exim now panics and dies. Previously, because the count
373 is held in a short int, deliveries were likely to be lost. As such a
374 large number of recipients for a single message is ridiculous
375 (performance will be very, very poor), I have chosen to impose a limit
376 rather than extend the field.
382 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
383 leftover -K file (the existence of which was triggered by #402).
384 While we were at it, introduced process PID as part of the -K
385 filename. This should rule out race conditions when creating
388 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
389 processing considerably. Previous code took too long for large mails,
390 triggering a timeout which in turn triggers #401.
392 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
393 in the DK code in transports.c. sendfile() is not really portable,
394 hence the _LINUX specificness.
396 TF/01 In the add_headers option to the mail command in an Exim filter,
397 there was a bug that Exim would claim a syntax error in any
398 header after the first one which had an odd number of characters
401 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
402 callout verification, Exim cached a "reject" for the entire domain. This
403 is correct for most verifications, but it is not correct for a recipient
404 verification with use_sender or use_postmaster set, because in that case
405 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
406 case of MAIL FROM:<> rejection from other early rejections (e.g.
407 rejection of HELO). When verifying a recipient using a non-null MAIL
408 address, the cache is ignored if it shows MAIL FROM:<> rejection.
409 Whatever the result of the callout, the value of the domain cache is
410 left unchanged (for any other kind of callout, getting as far as trying
411 RCPT means that the domain itself is ok).
413 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
416 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
417 manifest itself as EPIPE rather than ECONNECT. When tidying away a
418 session, the daemon ignores ECONNECT errors and logs others; it now
419 ignores EPIPE as well.
421 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
422 (quoted-printable decoding).
424 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
425 later the small subsequent patch to fix an introduced bug.
427 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
429 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
431 PH/08 An error is now given if message_size_limit is specified negative.
433 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
434 to be given (somewhat) arbitrary names.
436 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
439 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
440 miscellaneous code fixes
442 PH/10 Added the log_reject_target ACL modifier to specify where to log
445 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
446 hostname. This is wrong, because it relates to the incoming message (and
447 probably the interface on which it is arriving) and not to the outgoing
448 callout (which could be using a different interface). This has been
449 changed to use the value of the helo_data option from the smtp transport
450 instead - this is what is used when a message is actually being sent. If
451 there is no remote transport (possible with a router that sets up host
452 addresses), $smtp_active_hostname is used.
454 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
455 tweaks were necessary in order to get it to work (see also 21 below):
456 (a) The code assumed that strncpy() returns a negative number on buffer
457 overflow, which isn't the case. Replaced with Exim's string_format()
459 (b) There were several signed/unsigned issues. I just did the minimum
460 hacking in of casts. There is scope for a larger refactoring.
461 (c) The code used strcasecmp() which is not a standard C function.
462 Replaced with Exim's strcmpic() function.
463 (d) The code set only $1; it now sets $auth1 as well.
464 (e) A simple test gave the error "authentication client didn't specify
465 service in request". It would seem that Dovecot has changed its
466 interface. Fortunately there's a specification; I followed it and
467 changed what the client sends and it appears to be working now.
469 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
472 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
473 address A is aliased to B and C, where B exists and C does not. Without
474 -v the output is "A verified" because verification stops after a
475 successful redirection if more than one address is generated. However,
476 with -v the child addresses are also verified. Exim was outputting "A
477 failed to verify" and then showing the successful verification for C,
478 with its parentage. It now outputs "B failed to verify", showing B's
479 parentage before showing the successful verification of C.
481 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
482 look up a TXT record in a specific list after matching in a combined
485 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
486 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
487 they consult the DNS. I had assumed they would set it the way they
488 wanted; and indeed my experiments on Linux seem to show that in some
489 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
490 To be on the safe side, however, I have now made the interface to
491 host_find_byname() similar to host_find_bydns(), with an argument
492 containing the DNS resolver options. The host_find_byname() function now
493 sets these options at its start, just as host_find_bydns() does. The smtp
494 transport options dns_qualify_single and dns_search_parents are passed to
495 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
496 of host_find_byname() use the default settings of RES_DEFNAMES
497 (qualify_single) but not RES_DNSRCH (search_parents).
499 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
500 spool_read_header() do less string testing, by means of a preliminary
501 switch on the second character of optional "-foo" lines. (This is
502 overdue, caused by the large number of possibilities that now exist.
503 Originally there were few.) While I was there, I also converted the
504 str(n)cmp tests so they don't re-test the leading "-" and the first
505 character, in the hope this might squeeze out yet more improvement.
507 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
508 flag allowing group syntax was set by the header_syntax check but not
509 turned off, possible causing trouble later; (2) The flag was not being
510 set at all for the header_verify test, causing "group"-style headers to
511 be rejected. I have now set it in this case, and also caused header_
512 verify to ignore an empty address taken from a group. While doing this, I
513 came across some other cases where the code for allowing group syntax
514 while scanning a header line wasn't quite right (mostly, not resetting
515 the flag correctly in the right place). These bugs could have caused
516 trouble for malformed header lines. I hope it is now all correct.
518 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
519 with the "reply" argument non-NULL. The code, however (which originally
520 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
521 but it didn't always do it. This confused somebody who was copying the
522 code for some other use. I have removed all the tests.
524 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
525 feature that was used to support insecure browsers during the U.S. crypto
526 embargo. It requires special client support, and Exim is probably the
527 only MTA that supported it -- and would never use it because real RSA is
528 always available. This code has been removed, because it had the bad
529 effect of slowing Exim down by computing (never used) parameters for the
530 RSA_EXPORT functionality.
532 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
533 authenticator to fail if there's a tab character in the incoming data
534 (there should never be unless someone is messing about, as it's supposed
535 to be base64-encoded). Also added, on Timo's advice, the "secured" option
536 if the connection is using TLS or if the remote IP is the same as the
537 local IP, and the "valid-client-cert option" if a client certificate has
540 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
541 authenticators. This can be used for authorization after authentication
542 succeeds. (In the case of plaintext, it servers for both authentication
545 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
546 if any retry times were supplied.
548 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
549 connection, where there is no client IP address to check. In this
550 situation, the verify now always succeeds.
552 PH/25 Applied John Jetmore's -Mset patch.
554 PH/26 Added -bem to be like -Mset, but loading a message from a file.
556 PH/27 In a string expansion for a processed (not raw) header when multiple
557 headers of the same name were present, leading whitespace was being
558 removed from all of them, but trailing whitespace was being removed only
559 from the last one. Now trailing whitespace is removed from each header
560 before concatenation. Completely empty headers in a concatenation (as
563 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
564 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
566 PH/29 [Removed. This was a change that I later backed out, and forgot to
567 correct the ChangeLog entry (that I had efficiently created) before
568 committing the later change.]
570 PH/30 Exim was sometimes attempting to deliver messages that had suffered
571 address errors (4xx response to RCPT) over the same connection as other
572 messages routed to the same hosts. Such deliveries are always "forced",
573 so retry times are not inspected. This resulted in far too many retries
574 for the affected addresses. The effect occurred only when there were more
575 hosts than the hosts_max_try setting in the smtp transport when it had
576 the 4xx errors. Those hosts that it had tried were not added to the list
577 of hosts for which the message was waiting, so if all were tried, there
578 was no problem. Two fixes have been applied:
580 (i) If there are any address or message errors in an SMTP delivery, none
581 of the hosts (tried or untried) are now added to the list of hosts
582 for which the message is waiting, so the message should not be a
583 candidate for sending over the same connection that was used for a
584 successful delivery of some other message. This seems entirely
585 reasonable: after all the message is NOT "waiting for some host".
586 This is so "obvious" that I'm not sure why it wasn't done
587 previously. Hope I haven't missed anything, but it can't do any
588 harm, as the worst effect is to miss an optimization.
590 (ii) If, despite (i), such a delivery is accidentally attempted, the
591 routing retry time is respected, so at least it doesn't keep
592 hammering the server.
594 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
595 in ${readsocket because some servers need this prod.
597 PH/32 Added some extra debug output when updating a wait-xxx database.
599 PH/33 The hint "could be header name not terminated by colon", which has been
600 given for certain expansion errors for a long time, was not being given
601 for the ${if def:h_colon_omitted{... case.
603 PH/34 The spec says: "With one important exception, whenever a domain list is
604 being scanned, $domain contains the subject domain." There was at least
605 one case where this was not true.
607 PH/35 The error "getsockname() failed: connection reset by peer" was being
608 written to the panic log as well as the main log, but it isn't really
609 panic-worthy as it just means the connection died rather early on. I have
610 removed the panic log writing for the ECONNRESET error when getsockname()
613 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
614 runs only) independently of the message's sender address. This meant
615 that, if the 4xx error was in fact related to the sender, a different
616 message to the same recipient with a different sender could confuse
617 things. In particualar, this can happen when sending to a greylisting
618 server, but other circumstances could also provoke similar problems.
619 I have changed the default so that the retry time for these errors is now
620 based a combination of the sender and recipient addresses. This change
621 can be overridden by setting address_retry_include_sender=false in the
624 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
625 remote server are returned as part of bounce messages. This was not
626 happening for LMTP over a pipe (the lmtp transport), but now it is the
627 same for both kinds of LMTP.
629 PH/38 Despite being documented as not happening, Exim was rewriting addresses
630 in header lines that were in fact CNAMEs. This is no longer the case.
632 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
633 and queue runs started by the daemon processed all messages. This has
634 been fixed so that -R and -S can now usefully be given with -q<time>.
636 PH/40 Import PCRE release 6.7 (fixes some bugs).
638 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
640 PH/42 Give an error if -q is specified more than once.
642 PH/43 Renamed the variables $interface_address and $interface_port as
643 $received_ip_address and $received_port, to make it clear that these
644 values apply to message reception, and not to the outgoing interface when
645 a message is delivered. (The old names remain recognized, of course.)
647 PH/44 There was no timeout on the connect() call when using a Unix domain
648 socket in the ${readsocket expansion. There now is.
650 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
651 be meaningful with "accept".
653 SC/01 Eximstats V1.43
654 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
656 SC/02 Eximstats V1.44
657 Use a glob alias rather than an array ref in the generated
658 parser. This improves both readability and performance.
660 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
661 Collect SpamAssassin and rejection statistics.
662 Don't display local sender or destination tables unless
663 there is data to show.
664 Added average volumes into the top table text output.
666 SC/04 Eximstats V1.46
667 Collect data on the number of addresses (recipients)
668 as well as the number of messages.
670 SC/05 Eximstats V1.47
671 Added 'Message too big' to the list of mail rejection
672 reasons (thanks to Marco Gaiarin).
674 SC/06 Eximstats V1.48
675 Mainlog lines which have GMT offsets and are too short to
676 have a flag are now skipped.
678 SC/07 Eximstats V1.49 (Alain Williams)
679 Added the -emptyok flag.
681 SC/08 Eximstats V1.50
682 Fixes for obtaining the IP address from reject messages.
684 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
685 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
686 whitesspace changes from 4.64-PH/27
688 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
691 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
692 are found, allow negative numbers in numeric criteria)
694 JJ/06 exipick.20061117.2, added new $message_body_missing variable
696 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
697 to match changes made in 4.64-PH/43
699 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
701 PH/47 Put in an explicit test for a DNS lookup of an address record where the
702 "domain" is actually an IP address, and force a failure. This locks out
703 those revolvers/nameservers that support "A-for-A" lookups, in
704 contravention of the specifications.
706 PH/48 When a host name was looked up from an IP address, and the subsequent
707 forward lookup of the name timed out, the host name was left in
708 $sender_host_name, contrary to the specification.
710 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
711 restricted to single-key lookups, Exim was not diagnosing an error if
712 * or *@ was used with a query-style lookup.
714 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
716 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
717 long ago, but noone interested enough thought of it. Let's just say that
718 the "1.1" means that there are some new functions that weren't there at
719 some point in the past.
721 PH/51 Error processing for expansion failure of helo_data from an smtp
722 transport during callout processing was broken.
724 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
725 tested/used via the -bh/-bhc/-bs options.
727 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
728 bug, fixed in subsequent PCRE releases).
730 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
731 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
733 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
739 SC/01 Use a glob alias rather than an array ref in eximstats generated
740 parser. This improves both readability and performance.
742 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
743 Don't display local sender or destination tables in eximstats unless
744 there is data to show.
745 Added average volumes into the eximstats top table text output.
747 SC/03 Collect data on the number of addresses (recipients) as well
748 as the number of messages in eximstats.
750 TF/01 Correct an error in the documentation for the redirect router. Exim
751 does (usually) call initgroups() when daemonizing.
753 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
754 with consistent privilege compared to when running as a daemon.
756 TF/03 Note in the spec that $authenticated_id is not set for local
757 submissions from trusted users.
759 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
760 Thanks to Dean Brooks <dean@iglou.com> for the patch.
762 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
763 by adding some example configuration directives to the default
764 configuration file. A little bit of work is required to uncomment the
765 directives and define how usernames and passwords are checked, but
766 there is now a framework to start from.
768 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
769 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
770 without this. I don't know how relevant this is to other LDAP libraries.
772 PH/02 Add the verb name to the "unknown ACL verb" error.
774 PH/03 Magnus Holmgren's patch for filter_prepend_home.
776 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
778 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
779 directory not expanded when it should be if an expanded home directory
780 was set for the address (which is overridden by the transport).
782 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
785 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
786 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
787 because it is too late at that time, and has no effect.
789 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
790 security issue with \' (bugzilla #107). I could not use the
791 PQescapeStringConn() function, because it needs a PGconn value as one of
794 PH/08 When testing addresses using -bt, indicate those final addresses that
795 are duplicates that would not cause an additional delivery. At least one
796 person was confused, thinking that -bt output corresponded to deliveries.
797 (Suppressing duplicates isn't a good idea as you lose the information
798 about possibly different redirections that led to the duplicates.)
800 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
801 systems where poll() doesn't work, in particular OS X.
803 PH/10 Added more information to debugging output for retry time not reached.
805 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
806 operations in malware.c.
808 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
811 PH/13 If write_rejectlog was set false when logging was sent to syslog with
812 syslog_duplication set false, log lines that would normally be written
813 both the the main log and to the reject log were not written to syslog at
816 PH/14 In the default configuration, change the use of "message" in ACL warn
817 statements to "add_header".
819 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
820 not followed by a command (e.g. "seen endif").
822 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
823 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
826 PH/17 Added extra conditions to the default value of delay_warning_condition
830 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
831 { match{$h_precedence:}{(?i)bulk|list|junk} } \
832 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
835 The Auto-Submitted: and various List- headers are standardised, whereas I
836 don't think Precedence: ever was.
838 PH/18 Refactored debugging code in route_finduser() to show more information,
839 in particular, the error code if getpwnam() issues one.
841 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
842 This is apparently needed in addition to the PH/07 change above to avoid
843 any possible encoding problems.
845 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
846 but not after initializing Perl.
848 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
849 output them only if debugging. By default they are written stderr,
850 apparently, which is not desirable.
852 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
855 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
858 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
860 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
861 authenticated or an ident call has been made. Suppress the default
862 values for $authenticated_id and $authenticated_sender (but permit -oMai
863 and -oMas) when testing with -bh.
865 PH/24 Re-jigged the order of the tests in the default configuration so that the
866 tests for valid domains and recipients precede the DNS black list and CSA
867 tests, on the grounds that those ones are more expensive.
869 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
870 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
871 This bug exists in every version of Exim that I still have, right back to
874 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
875 However, an attempt to turn on case-sensitivity in a regex key by
876 including (?-i) didn't work because the subject string was already
877 lowercased, and the effects were non-intuitive. It turns out that a
878 one-line patch can be used to allow (?-i) to work as expected.
884 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
885 other effects) broke the use of negated acl sub-conditions.
887 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
890 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
891 "Deny" causes Exim to reject the incoming connection with a 554 error.
892 Unfortunately, if there is a major crisis, such as a disk failure,
893 tcp-wrappers gives "deny", whereas what one would like would be some
894 kind of temporary error. A kludge has been added to help with this.
895 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
896 554 error is used if errno is still zero or contains ENOENT (which occurs
897 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
900 PH/03 Add -lutil to the default FreeBSD LIBS setting.
902 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
903 errors. Otherwise a message that provokes a temporary error (when other
904 messages do not) can cause a whole host to time out.
906 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
907 addresses were routed directly to files or pipes from a redirect router.
908 File deliveries just didn't batch; pipe deliveries might have suffered
911 PH/06 A failure to get a lock for a hints database would erroneously always say
912 "Failed to get write lock", even when it was really a read lock.
914 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
915 of 0600. This has been changed to use the value of the lockfile_mode
916 option (which defaults to 0600).
918 PH/08 Applied small patch from the Sieve maintainer.
920 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
921 folder from quota calculations, a direct delivery into this folder messed
922 up the contents of the maildirsize file. This was because the regex was
923 used only to exclude .Trash (or whatever) when the size of the mailbox
924 was calculated. There was no check that a delivery was happening into an
925 excluded directory. This bug has been fixed by ignoring all quota
926 processing for deliveries into excluded directories.
928 PH/10 Added the maildirfolder_create_regex option to appendfile.
934 PH/01 The code for finding all the local interface addresses on a FreeBSD
935 system running IPv6 was broken. This may well have applied to all BSD
936 systems, as well as to others that have similar system calls. The broken
937 code found IPv4 interfaces correctly, but gave incorrect values for the
938 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
939 that it would not match correctly against @[] and not recognize the IPv6
942 PH/02 The ipliteral router was not recognizing addresses of the form user@
943 [ipv6:....] because it didn't know about the "ipv6:" prefix.
945 PH/03 Added disable_ipv6.
947 PH/04 Changed $reply_address to use the raw form of the headers instead of the
948 decoded form, because it is most often used to construct To: headers
949 lines in autoreplies, and the decoded form may well be syntactically
950 invalid. However, $reply_address has leading white space removed, and all
951 newlines turned into spaces so that the autoreply transport does not
954 PH/05 If group was specified without a user on a router, and no group or user
955 was specified on a transport, the group from the router was ignored.
957 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
958 for visible compile-time settings that can be used to change these
959 numbers, for those that want even more. Backwards compatibility with old
960 spool files has been maintained. However, going back to a previous Exim
961 release will lost any variables that are in spool files.
963 PH/07 Two small changes when running in the test harness: increase delay when
964 passing a TCP/IP connection to a new process, in case the original
965 process has to generate a bounce, and remove special handling of
966 127.0.0.2 (sic), which is no longer necessary.
968 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
969 be the same on different OS.
971 PH/09 Moved a debug statement in filter processing to avoid a race problem when
974 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
975 whether --show-vars was specified or not
977 JJ/02 exipick: Added support for new ACL variable spool format introduced
980 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
981 syntactically invalid From: or Reply-to: line, and a filter used this to
982 generate an autoreply, and therefore failed to obtain an address for the
983 autoreply, Exim could try to deliver to a non-existent relative file
984 name, causing unrelated and misleading errors. What now happens is that
985 it logs this as a hard delivery error, but does not attempt to create a
988 PH/11 The exinext utility has a -C option for testing purposes, but although
989 the given file was scanned by exinext itself; it wasn't being passed on
990 when Exim was called.
992 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
993 an end-of-file indication when reading a command response.
995 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
996 compiled. In many other places in Exim, IPv6 addresses are always
997 recognized, so I have changed this. It also means that IPv4 domain
998 literals of the form [IPV4:n.n.n.n] are now always recognized.
1000 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1001 used if the router is not running as root, for example, when verifying at
1002 ACL time, or when using -bh. The debugging output from this situation was
1003 non-existent - all you got was a failure to exec. I have made two
1006 (a) Failures to set uid/gid, the current directory, or a process leader
1007 in a subprocess such as that created by queryprogram now generate
1008 suitable debugging ouput when -d is set.
1010 (b) The queryprogram router detects when it is not running as root,
1011 outputs suitable debugging information if -d is set, and then runs
1012 the subprocess without attempting to change uid/gid.
1014 PH/15 Minor change to Makefile for building test_host (undocumented testing
1017 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1018 additional section of a DNS packet that returns MX or SRV records.
1019 Instead, it always explicitly searches for A/AAAA records. This avoids
1020 major problems that occur when a DNS server includes only records of one
1021 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1022 fixed another bug: if SRV records were looked up and the corresponding
1023 address records were *not* found in the additional section, the port
1024 values from the SRV records were lost.
1026 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1027 using the correct key (the original address) when searching the retry
1028 rules in order to find which one to use for generating the retry hint.
1030 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1031 adding the default one. Similarly, if it contains a Reply-To: header, the
1032 errors_reply_to option, if set, is not used.
1034 PH/19 When calculating a retry time, Exim used to measure the "time since
1035 failure" by looking at the "first failed" field in the retry record. Now
1036 it does not use this if it is later than than the arrival time of the
1037 message. Instead it uses the arrival time. This makes for better
1038 behaviour in cases where some deliveries succeed, thus re-setting the
1039 "first failed" field. An example is a quota failure for a huge message
1040 when small messages continue to be delivered. Without this change, the
1041 "time since failure" will always be short, possible causing more frequent
1042 delivery attempts for the huge message than are intended.
1043 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1045 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1046 $1, $2, $3) because the numerical variables can be reset during some
1047 expansion items (e.g. "match"), thereby losing the authentication data.
1049 PH/21 Make -bV show the size of off_t variables so that the test suite can
1050 decide whether to run tests for quotas > 2G.
1052 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1053 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1054 filecount value is greater than 2G or if a quota value is greater than 2G
1055 on a system where the size of off_t is not greater than 4, a panic error
1058 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1059 never match. The debug and -bh output now contains an explicit error
1060 message indicating a malformed IPv4 address or mask.
1062 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1063 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1064 PH/23 above applies.
1066 PH/25 Do not write to syslog when running in the test harness. The only
1067 occasion when this arises is a failure to open the main or panic logs
1068 (for which there is an explicit test).
1070 PH/26 Added the /no_tell option to "control=freeze".
1072 PH/27 If a host name lookup failed very early in a connection, for example, if
1073 the IP address matched host_lookup and the reverse lookup yielded a name
1074 that did not have a forward lookup, an error message of the form "no IP
1075 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1076 could be logged. Now it outputs the IP address instead of "NULL".
1078 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1079 allows the sender and the authenticated sender to be set when
1080 submitting a message from within Exim. Since child_open_exim() is
1081 documented for local_scan(), the new function should be too.
1083 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1084 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1085 results in an empty string is now treated as unset.
1087 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1089 PH/31 Added sender_verify_fail logging option.
1091 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1092 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1093 tidied the source and removed it altogether.
1095 PH/33 When a queue run was abandoned because the load average was too high, a
1096 log line was always written; now it is written only if the queue_run log
1097 selector is set. In addition, the log line for abandonment now contains
1098 information about the queue run such as the pid. This is always present
1099 in "start" and "stop" lines but was omitted from the "abandon" line.
1101 PH/34 Omit spaces between a header name and the colon in the error message that
1102 is given when verify = headers_syntax fails (if there are lots of them,
1103 the message gets confusing).
1105 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1106 names, as there are now some PTR records that contain slashes. This check
1107 is only to protect against broken name servers that fall over on strange
1108 characters, so the fact that it applies to all lookups doesn't matter.
1110 PH/36 Now that the new test suite is complete, we can remove some of the
1111 special code in Exim that was needed for the old test suite. For example,
1112 sorting DNS records because real resolvers return them in an arbitrary
1113 order. The new test suite's fake resolver always returns records in the
1116 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1117 bounces) except when queue_only is set, to avoid logging races between
1118 the different processes.
1120 PH/38 Panic-die if .include specifies a non-absolute path.
1122 PH/39 A tweak to the "H" retry rule from its user.
1124 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1125 a label. They prevented compilation on older perls.
1127 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1128 a warning to be raised on newish perls.
1130 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1131 on queue. Changes to match documented behaviour of showing count of
1132 messages matching specified criteria.
1134 PH/40 Changed the default ident timeout from 30s to 5s.
1136 PH/41 Added support for the use of login_cap features, on those BSD systems
1137 that have them, for controlling the resources used by pipe deliveries.
1139 PH/42 The content-scanning code uses fopen() to create files in which to put
1140 message data. Previously it was not paying any attention to the mode of
1141 the files. Exim runs with umask(0) because the rest of the code creates
1142 files with open(), and sets the required mode explicitly. Thus, these
1143 files were ending up world-writeable. This was not a big issue, because,
1144 being within the spool directory, they were not world-accessible. I have
1145 created a function called modefopen, which takes an additional mode
1146 argument. It sets umask(777), creates the file, chmods it to the required
1147 mode, then resets the umask. All the relevant calls to fopen() in the
1148 content scanning code have been changed to use this function.
1150 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1151 to 24 hours. This avoids potential overflow problems when processing G
1152 and H retry rules. I suspect nobody ever tinkers with this value.
1154 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1156 PH/45 When the plaintext authenticator is running as a client, the server's
1157 challenges are checked to ensure they are valid base64 strings. By
1158 default, the authentication attempt is cancelled if an invalid string is
1159 received. Setting client_ignore_invalid_base64 true ignores these errors.
1160 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1161 they are received. Thus, the responses can be made to depend on the
1162 challenges. If an invalid string is ignored, an empty string is placed in
1165 PH/46 Messages that are created by the autoreply transport now contains a
1166 References: header, in accordance with RFCs 2822 and 3834.
1168 PH/47 Added authenticated_sender_force to the smtp transport.
1170 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1172 PH/49 Installed latest patch from the Sieve maintainer.
1174 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1175 was also set, the appendfile transport was unnecessarily scanning a
1176 directory of message files (e.g. for maildir delivery) to find the count
1177 of files (along with the size), even though it did not need this
1178 information. It now does the scan only if it needs to find either the
1179 size of the count of files.
1181 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1183 PH/52 Two bugs concerned with error handling when the smtp transport is
1186 (i) Exim was not creating retry information for temporary errors given
1187 for individual recipients after the DATA command when the smtp transport
1188 was used in LMTP mode. This meant that they could be retried too
1189 frequently, and not timed out correctly.
1191 (ii) Exim was setting the flag that allows error details to be returned
1192 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1193 recipients that were returned after the DATA command.
1195 PH/53 This is related to PH/52, but is more general: for any failing address,
1196 when detailed error information was permitted to be returned to the
1197 sender, but the error was temporary, then after the final timeout, only
1198 "retry timeout exceeded" was returned. Now it returns the full error as
1199 well as "retry timeout exceeded".
1201 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1202 do this, and (what is worse) MTAs that accept it.
1204 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1205 will now be deprecated.
1207 PH/56 New os.c-cygwin from the Cygwin maintainer.
1209 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1210 formats (previously only available in exim formats via -bpr, -bpru,
1211 and -bpra. Now also available in native and exiqgrep formats)
1213 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1214 with very large, slow to parse queues
1216 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1218 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1220 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1221 responses to authentication challenges, though it was showing the
1222 challenges; (ii) I've removed the CR characters from the debug output for
1225 PH/58 Allow for the insertion of a newline as well as a space when a string
1226 is turned into more than one encoded-word during RFC 2047 encoding. The
1227 Sieve code now uses this.
1229 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1230 data_4xx, lost_connection, tls_required.
1232 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1233 message was being sent as an SMTP response.
1235 PH/61 Add -l and -k options to exicyclog.
1237 PH/62 When verifying, if an address was redirected to one new address, so that
1238 verification continued, and the new address failed or deferred after
1239 having set something in $address_data, the value of $address_data was not
1240 passed back to the ACL. This was different to the case when no
1241 redirection occurred. The value is now passed back in both cases.
1243 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1244 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1245 use login_cap.h, so on its own it isn't the distinguishing feature. The
1246 new name refers directly to the setclassresources() function.
1248 PH/65 Added configuration files for NetBSD3.
1250 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1252 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1253 is preferred over IPv4.
1255 PH/68 The bounce_return_message and bounce_return_body options were not being
1256 honoured for bounces generated during the reception of non-SMTP messages.
1257 In particular, this applied to messages rejected by the ACL. This bug has
1258 been fixed. However, if bounce_return_message is true and bounce_return_
1259 body is false, the headers that are returned for a non-SMTP message
1260 include only those that have been read before the error was detected.
1261 (In the case of an ACL rejection, they have all been read.)
1263 PH/69 The HTML version of the specification is now built in a directory called
1264 spec_html instead of spec.html, because the latter looks like a path with
1265 a MIME-type, and this confuses some software.
1267 PH/70 Catch two compiler warnings in sieve.c.
1269 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1270 function verify_get_ident() calls ip_connect() to connect a socket, but
1271 if the "connect()" function timed out, ip_connect() used to close the
1272 socket. However, verify_get_ident() also closes the socket later, and in
1273 between Exim writes to the log, which may get opened at this point. When
1274 the socket was closed in ip_connect(), the log could get the same file
1275 descriptor number as the socket. This naturally causes chaos. The fix is
1276 not to close the socket in ip_connect(); the socket should be closed by
1277 the function that creates it. There was only one place in the code where
1278 this was missing, in the iplookup router, which I don't think anybody now
1279 uses, but I've fixed it anyway.
1281 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1282 well as to direct DNS lookups. Otherwise the handling of names in host
1283 lists is inconsistent and therefore confusing.
1289 PH/01 Two changes to the default runtime configuration:
1291 (1) Move the checks for relay_from_hosts and authenticated clients from
1292 after to before the (commented out) DNS black list checks.
1294 (2) Add control=submission to the relay_from_hosts and authenticated
1295 clients checks, on the grounds that messages accepted by these
1296 statements are most likely to be submissions.
1298 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1300 (1) Generate an error if the third argument for the ${prvs expansion is
1303 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1306 (3) Reset the variables that are obtained from the first argument of
1307 ${prvscheck and used in the second argument before leaving the code,
1308 because their memory is reclaimed, so using them afterwards may do
1311 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1312 one (it's much easier than Tom thought :-).
1314 (5) Because of (4), we can now allow for the use of $prvscheck_result
1315 inside the third argument.
1317 PH/03 For some reason, the default setting of PATH when running a command from
1318 a pipe transport was just "/usr/bin". I have changed it to
1321 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1322 anything to be listed in the output from -bV.
1324 PH/05 When a filter generated an autoreply, the entire To: header line was
1325 quoted in the delivery log line, like this:
1327 => >A.N.Other <ano@some.domain> <original@ddress> ...
1329 This has been changed so that it extracts the operative address. There
1330 may be more than one such address. If so, they are comma-separated, like
1333 => >ano@some.domain,ona@other.domain <original@ddress> ...
1335 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1336 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1337 not being looked up in the rDNS to get a host name, Exim was showing the
1338 IP address twice in Received: lines, even though the IP addresses were
1339 identical. For example:
1341 Received: from [1.2.3.4] (helo=[1.2.3.4])
1343 However, if the real host name was known, it was omitting the HELO data
1344 if it matched the actual IP address. This has been tidied up so that it
1345 doesn't show the same IP address twice.
1347 PH/07 When both +timestamp and +memory debugging was on, the value given by
1348 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1349 called by the debug printing, thereby overwriting the timestamp buffer.
1350 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1352 PH/08 When the original message was included in an autoreply transport, it
1353 always said "this is a copy of the message, including all the headers",
1354 even if body_only or headers_only was set. It now gives an appropriate
1357 PH/09 Applied a patch from the Sieve maintainer which:
1359 o fixes some comments
1360 o adds the (disabled) notify extension core
1361 o adds some debug output for the result of if/elsif tests
1362 o points to the current vacation draft in the documentation
1363 and documents the missing references header update
1367 o fixes a bug in processing the envelope test (when testing
1368 multiple envelope elements, the last element determinted the
1371 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1372 Electronic Mail") by including:
1374 Auto-submitted: auto-generated
1376 in the messages that it generates (bounce messages and others, such as
1377 warnings). In the case of bounce messages for non-SMTP mesages, there was
1378 also a typo: it was using "Auto_submitted" (underscore instead of
1379 hyphen). Since every message generated by Exim is necessarily in response
1380 to another message, thes have all been changed to:
1382 Auto-Submitted: auto-replied
1384 in accordance with these statements in the RFC:
1386 The auto-replied keyword:
1388 - SHOULD be used on messages sent in direct response to another
1389 message by an automatic process,
1391 - MUST NOT be used on manually-generated messages,
1393 - MAY be used on Delivery Status Notifications (DSNs) and Message
1394 Disposition Notifications (MDNs),
1396 - MUST NOT be used on messages generated by automatic or periodic
1397 processes, except for messages which are automatic responses to
1400 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1401 to the default Received: header definition.
1403 PH/12 Added log selector acl_warn_skipped (default on).
1405 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1406 variables because (a) they are in the wrong storage pool and (b) even if
1407 they were copied, it wouldn't work properly because of the caching.
1409 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1410 checking when decoding. Apparently there are clients that generate
1411 overlong encoded strings. Why am I not surprised?
1413 PH/15 If the first argument of "${if match_address" was not empty, but did not
1414 contain an "@" character, Exim crashed. Now it writes a panic log message
1415 and treats the condition as false.
1417 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1419 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1420 extension "envelope-auth". The code is finished and in agreement with
1421 other implementations, but there is no documentation so far and in fact,
1422 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1423 not changing the active code.
1425 Print executed "if" and "elsif" statements when debugging is used. This
1426 helps a great deal to understand what a filter does.
1428 Document more things not specified clearly in RFC3028. I had all this
1429 sorted out, when out of a sudden new issues came to my mind. Oops."
1431 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1434 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1435 canonical form (as documented). However, after a host name lookup from
1436 the IP address, check_host() was doing a simple string comparison with
1437 addresses acquired from the DNS when checking that the found name did
1438 have the original IP as one of its addresses. Since any found IPv6
1439 addresses are likely to be in abbreviated form, the comparison could
1440 fail. Luckily, there already exists a function for doing the comparison
1441 by converting both addresses to binary, so now that is used instead of
1442 the text comparison.
1444 PH/20 There was another similar case to PH/19, when a complete host name was
1445 given in a host list; looking up its IP address could give an abbreviated
1446 form, whereas the current host's name might or might not be abbreviated.
1447 The same fix has been applied.
1453 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1454 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1457 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1458 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1460 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1462 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1463 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1464 Also, for "Auto-Submitted", treat anything other than "no" as
1465 non-personal, in accordance with RFC 3834. (Previously it treated
1466 anything starting "auto-" as non-personal.)
1468 TF/01 The control=submission/name=... option had a problem with syntax
1469 errors if the name included a slash character. The /name= option
1470 now slurps the rest of the string, so it can include any characters
1471 but it must come last in the list of options (after /sender_retain
1474 PH/05 Some modifications to the interface to the fake nameserver for the new
1482 TK/01 Added the "success_on_redirect" address verification option. See
1483 NewStuff for rationale and an example.
1485 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1487 PH/02 Patch to exigrep to allow it to work on syslog lines.
1489 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1490 fread() to skip over the body file's header line, because in Cygwin the
1491 header line is locked and is inaccessible.
1493 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1494 co-exist for some time) to make it clear that it is the Exim ID that is
1495 referenced, not the Message-ID: header line.
1497 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1498 string_format() function, because snprintf() does not exist on all
1501 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1502 ${stat: expansion item.
1504 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1505 protocol synchronization error", to keep the pedants happy.
1507 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1508 well as for IRIX systems, when gcc is being used. See the host.c source
1511 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1513 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1516 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1517 transport and to the smtp transport in LMTP mode.
1519 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1521 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1522 run a filter in a subprocess. This could lead to confusion in subsequent
1523 lookups in the parent process. There should also be a search_tidyup() at
1524 the end of the subprocess.
1526 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1527 only if the host matched helo_try_verify_hosts, which caused the
1528 verification to occur when the EHLO/HELO command was issued. The ACL just
1529 tested the remembered result. Now, if a previous verification attempt has
1530 not happened, "verify = helo" does it there and then.
1532 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1534 TK/03 Fix log output including CR from clamd.
1536 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1537 exist provoked a memory error which could cause a segfault.
1539 PH/15 Installed PCRE 6.2
1541 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1543 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1544 of the problem. Specifically, suggested +O2 rather than +O1 for the
1547 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1549 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1550 record, the port was not correctly logged when the outgoing_port log
1551 selector was set (it logged the transort's default port).
1553 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1554 fallback_hosts, and "hosts" in the smtp transport.
1556 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1557 host errors such as "Connection refused".
1559 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1560 authentication with radiusclient 0.4.9:
1562 - Error returned from rc_read_config was caught wrongly
1563 - Username/password not passed on to radius server due to wrong length.
1565 The presumption is that some radiusclient API changes for 4.51/PH/17
1566 were not taken care of correctly. The code is still untested by me (my
1567 Linux distribution still has 0.3.2 of radiusclient), but it was
1568 contributed by a Radius user.
1570 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1571 expanding the "port" option of the smtp transport.
1573 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1574 while reading a MIME header. Thanks to Tom Hughes for a patch.
1576 PH/24 Include config.h inside local_scan.h so that configuration settings are
1579 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1580 that in an ACL for RCPT (for example), you can examine exactly what was
1583 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1584 commands, but it was not correctly comparing the address with the actual
1585 client host address. Thus, it would show the EHLO address in Received:
1586 header lines when this was not necessary.
1588 PH/27 Added the % operator to ${eval:}.
1590 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1591 it should be ignoring failures (because with -C, for example, it has lost
1592 privilege). It wasn't ignoring creation failures other than "already
1595 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1598 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1599 because an input line was too long, either on its own, or by virtue of
1600 too many continuations, the temporary file was not being removed, and the
1601 return code was incorrect.
1603 PH/31 Missing "BOOL" in function definition in filtertest.c.
1605 PH/32 Applied Sieve patches from the maintainer.
1607 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1609 PH/33 Added "verify = not_blind".
1611 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1612 Local/Makefile (with some defaults set). These are used in built scripts
1613 such as exicyclog, but they have never been used in the exim_install
1614 script (though there are many overriding facilities there). I have
1615 arranged that the exim_install script now takes note of these two
1618 PH/35 Installed configuration files for Dragonfly.
1620 PH/36 When a locally submitted message by a trusted user did not contain a
1621 From: header, and the sender address was obtained from -f or from an SMTP
1622 MAIL command, and the trusted user did not use -F to supply a sender
1623 name, $originator_name was incorrectly used when constructing a From:
1624 header. Furthermore, $originator_name was used for submission mode
1625 messages from external hosts without From: headers in a similar way,
1626 which is clearly wrong.
1628 PH/37 Added control=suppress_local_fixups.
1630 PH/38 When log_selector = +received_sender was set, and the addition of the
1631 sender made the log line's construction buffer exactly full, or one byte
1632 less than full, an overflow happened when the terminating "\n" was
1635 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1636 when the result of a list match is failure because a DNS lookup failed.
1638 PH/40 RM_COMMAND is now used in the building process.
1640 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1641 the "build-* directories that it finds.
1643 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1644 address in a domain literal was a prefix of an interface address.
1646 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1647 when verifying a sender address, unless rewrite_headers is false.
1649 PH/44 Wrote a long comment about why errors_to addresses are verified as
1650 recipients, not senders.
1652 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1653 the ratelimit ACL was added.
1655 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1657 PH/46 Added extra information about PostgreSQL errors to the error string.
1659 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1660 avoiding the need to install special zones in a real server. This is
1661 backwards compatible; if it can't find the fake resolver, it drops back.
1662 Thus, both old and new test suites can be run.
1664 TF/02 Added util/ratelimit.pl
1666 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1667 clock is set back in time.
1669 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1670 Candler <B.Candler@pobox.com>.
1672 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1673 OK for addresses that are the result of redirections.
1675 PH/48 A number of further additions for the benefit of the new test suite,
1676 including a fake gethostbyname() that interfaces to the fake DNS resolver
1679 TF/06 The fix for widen_domains has also been applied to qualify_single and
1680 search_parents which are the other dnslookup options that can cause
1683 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1686 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1688 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1689 service" if it encountered an MX record with an empty target hostname.
1690 The message is now "an MX or SRV record indicated no SMTP service".
1692 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1693 if the DNS of the sending site is misconfigured. This is quite a
1694 common situation. This change restores the behaviour of treating a
1695 helo verification defer as a failure.
1697 PH/51 If self=fail was set on a router, the bounce message did not include the
1698 actual error message.
1704 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1706 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1707 command itself ended in error, the underlying message about the transport
1708 filter timeout was being overwritten with the pipe command error. Now the
1709 underlying error message should be appended to the second error message.
1711 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1713 PH/02 Reduce the amount of output that "make" produces by default. Full output
1714 can still be requested.
1716 PH/03 The warning log line about a condition test deferring for a "warn" verb
1717 was being output only once per connection, rather than after each
1718 occurrence (because it was using the same function as for successful
1719 "warn" verbs). This seems wrong, so I have changed it.
1721 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1722 it should not have, which might have caused a crash in the right
1723 circumstances, but probably never did.
1725 PH/04 Installed a modified version of Tony Finch's patch to make submission
1726 mode fix the return path as well as the Sender: header line, and to
1727 add a /name= option so that you can make the user's friendly name appear
1730 TF/03 Added the control = fakedefer ACL modifier.
1732 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1733 Mark Lowes for thorough testing.
1735 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1737 TK/03 Merged latest SRS patch from Miles Wilton.
1739 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1740 with the definition in sysexits.h (which is #included earlier).
1741 Fortunately, Exim does not actually use EX_OK. The code used to try to
1742 preserve the sysexits.h value, by assumimg that macro definitions were
1743 scanned for macro replacements. I have been disabused of this notion,
1744 so now the code just undefines EX_OK before #including unistd.h.
1746 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1747 in the smtp transport. When a block could not be written in a single
1748 write() function, the timeout was being re-applied to each part-write.
1749 This seems wrong - if the receiver was accepting one byte at a time it
1750 would take for ever. The timeout is now adjusted when this happens. It
1751 doesn't have to be particularly precise.
1753 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1754 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1756 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1757 without a domain if the check to <postmaster@domain> fails.
1759 SC/01 Eximstats: added -xls and the ability to specify output files
1760 (patch written by Frank Heydlauf).
1762 SC/02 Eximstats: use FileHandles for outputing results.
1764 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1766 SC/04 Eximstats: fixed display of large numbers with -nvr option
1768 SC/05 Eximstats: fixed merging of reports with empty tables.
1770 SC/06 Eximstats: added the -include_original_destination flag
1772 SC/07 Eximstats: removed tabs and trailing whitespace.
1774 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1776 TK/06 MBOX spool code: Add real "From " MBOX separator line
1777 so the .eml file is really in mbox format (even though
1778 most programs do not really care). Patch from Alex Miller.
1780 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1781 The latter is generated from $received_to and is only set if the
1782 message has one envelope recipient. SA can use these headers,
1783 obviously out-of-the-box. Patch from Alex Miller.
1785 PH/08 The ${def test on a variable was returning false if the variable's
1786 value was "0", contrary to what the specification has always said!
1787 The result should be true unless the variable is empty.
1789 PH/09 The syntax error of a character other than { following "${if
1790 def:variable_name" (after optional whitespace) was not being diagnosed.
1791 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1792 accidental colon was present, for example, could give incorrect results.
1794 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1795 result is used (not including appendfile, where other changes are about
1798 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1799 This involved changing a lot of size variables from int to off_t. It
1800 should work with maildirs and everything.
1802 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1803 spamd dying while we are connected to it.
1805 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1808 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1809 function for BDB 4.3.
1811 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1813 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1814 the sources for PCRE have been re-arranged and more files are now
1817 PH/15 The code I had for printing potentially long long variables in PH/11
1818 above was not the best (it lost precision). The length of off_t variables
1819 is now inspected at build time, and an appropriate printing format (%ld
1820 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1821 to be "long long int" or "long int". This is needed for the internal
1822 formatting function string_vformat().
1824 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1825 the configuration file to be ":syslog", then the script "guesses" where
1826 the logs files are, rather than using the compiled in default. In our
1827 case the guess is not the same as the compiled default, so the script
1828 suddenly stopped working when I started to use syslog. The patch checks
1829 to see if log_file_path is "". If so, it attempts to read it from exim
1830 with no configuration file to get the compiled in version, before it
1831 falls back to the previous guessing code."
1833 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1834 implementing BATV in an Exim configuration. See NewStuff for the gory
1837 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1838 Makefile that are specific to HP-UX.
1840 PH/18 If the "use_postmaster" option was set for a recipient callout together
1841 with the "random" option, the postmaster address was used as the MAIL
1842 FROM address for the random test, but not for the subsequent recipient
1843 test. It is now used for both.
1845 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1846 patch removes a few documentation additions to RFC 3028, because the
1847 latest draft now contains them. It adds the new en;ascii-case comparator
1848 and a new error check for 8bit text in MIME parts. Comparator and
1849 require names are now matched exactly. I enabled the subaddress
1850 extension, but it is not well tested yet (read: it works for me)."
1852 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1853 rework some of the code of TK/09 above to avoid the hardwired use of
1854 "%lld" and "long long". Replaced the call to snprintf() with a call to
1857 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1858 records point to non-existent hosts", "retry timeout exceeded", and
1859 "retry time not reached for any host after a long failure period".
1861 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1862 experimental DomainKeys support:
1864 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1865 (2) On an error such as an illegally used "control", the wrong name for
1866 the control was given.
1868 These problems did NOT occur unless DomainKeys support was compiled.
1870 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1872 PH/24 Added ${if match_ip condition.
1874 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1875 databases so that it will be absolutely obvious if a crash occurs in the
1876 DB library. This is a regular occurrence (often caused by mis-matched
1879 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1880 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1881 picked up on a user's system that detects such things. There doesn't seem
1882 to be a gcc warning option for this - only an attribute that has to be
1883 put on the function's prototype. It seems that in Fedora Core 4 they have
1884 set this on a number of new functions. No doubt there will be more in due
1887 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1888 specify a transport. However, if an address that was verified by such a
1889 router was the subject of a callout, Exim crashed because it tried to
1890 read the rcpt_include_affixes from the non-existent transport. Now it
1891 just assumes that the setting of that option is false. This bug was
1892 introduced by 4.51/PH/31.
1894 PH/28 Changed -d+all to exclude +memory, because that information is very
1895 rarely of interest, but it makes the output a lot bigger. People tend to
1896 do -d+all out of habit.
1898 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1899 code in os-type was giving problems when libc.so lives in lib64, like on
1902 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1903 aren't the modern standard, and it seems that some systems' include files
1904 don't always have them. Exim was already checking for some of the newer
1905 ones like T_AAAA, and defining it itself. I've added checks for all the
1906 record types that Exim uses.
1908 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1909 not automatically generating a new one, as it is supposed to. This
1910 prevented TLS from working. If the file did exist, but contained invalid
1911 data, a new version was generated, as expected. It was only the case of a
1912 non-existent file that was broken.
1914 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1915 with a change in libdomainkeys > 0.64.
1917 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1918 from DNS. If the selector record carries the flag, it now has
1919 precedence over the domain-wide flag.
1921 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1923 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1924 the use of a port name that isn't defined in /etc/services) occurred, the
1925 message was deferred as in a normal delivery, and thus remained on the
1926 spool, instead of being failed because of the mua_wrapper setting. This
1927 is now fixed, and I tidied up some of the mua_wrapper messages at the
1930 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1931 the messages in a hash of arrays rather than using individual hashes.
1932 This is a bit cleaner and results in dramatic memory savings, albeit
1933 at a slight CPU cost.
1935 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1936 as requested by Marc Sherman.
1938 SC/10 Eximstats: added histograms for user specified patterns as requested
1941 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1943 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1944 fopen() in the content-scanning modules that did not already have it.
1950 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1951 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1953 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1955 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1957 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1958 to test Sieve filters that use "vacation".
1960 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1961 that changes the way the GnuTLS parameters are stored in the cache file.
1962 The new format can be generated externally. For backward compatibility,
1963 if the data in the cache doesn't make sense, Exim assumes it has read an
1964 old-format file, and it generates new data and writes a new file. This
1965 means that you can't go back to an older release without removing the
1968 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1969 work if there are any delivery delays because "one_time" forces the
1970 parent to be marked "delivered", so its unseen clone is never tried
1971 again. For this reason, Exim now forbids the simultaneous setting of
1974 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1975 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1976 entry if you want to know the details. The fix, however, neglected to
1977 consider the case when local delivery batching is involved. The test for
1978 "previously delivered" was not happening when checking to see if an
1979 address could be batched with a previous (undelivered) one; under
1980 certain circumstances this could lead to multiple deliveries to the same
1983 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1984 in its include files, and this causes problems building Exim.
1986 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1987 header_syntax) but Exim was just ignoring anything given after a slash.
1988 In particular, this caused confusion with an attempt to use "verify =
1989 reverse_host_lookup/defer_ok". An error is now given when options are
1990 supplied for verify items that do not have them. (Maybe reverse_host_
1991 lookup should have a defer_ok option, but that's a different point.)
1993 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1994 defined by RFC 821) to 2048, because there were problems with some AUTH
1995 commands, and RFC 1869 says the size should be increased for extended
1996 SMTP commands that take arguments.
1998 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2001 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2002 "unknown" error; now it says that the functionality isn't in the binary.
2004 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2005 an address' error message when a string expansion fails (syntax or
2006 whatever). Otherwise the password may appear in the log. Following change
2007 PH/42 below, there is no longer a chance of it appearing in a bounce
2010 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2012 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2013 of its addresses was ever tried. (Bugzilla bug #2.)
2015 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2016 the result incorrectly in the debug output. (It correctly added a newline
2017 to what was transported.)
2019 TF/01 Added $received_time.
2021 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2022 commented out examples of how to interface to a virus scanner and to
2023 SpamAssassin. Also added commented examples of av_scanner and
2024 spamd_address settings.
2026 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2027 and controls are allowed in which ACLs. There were a couple of minor
2028 errors. Some of the entries in the conditions table (which is a table of
2029 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2030 negation of where the condition IS allowed.
2032 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2034 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2035 header file does not have a version number, so I've had to invent a new
2036 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2037 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2038 radiusclient), but it was contributed by a Radius user.
2040 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2041 files or directories, for OpenSSL.
2043 PH/19 When an Exim process that is running as root has to create an Exim log
2044 file, it does so in a subprocess that runs as exim:exim so as to get the
2045 ownership right at creation (otherwise, other Exim processes might see
2046 the file with the wrong ownership). There was no test for failure of this
2047 fork() call, which would lead to the process getting stuck as it waited
2048 for a non-existent subprocess. Forks do occasionally fail when resources
2049 run out. I reviewed all the other calls to fork(); they all seem to check
2052 PH/20 When checking for unexpected SMTP input at connect time (before writing
2053 the banner), Exim was not dealing correctly with a non-positive return
2054 from the read() function. If the client had disconnected by this time,
2055 the result was a log entry for a synchronization error with an empty
2056 string after "input=" when read() returned zero. If read() returned -1
2057 (an event I could not check), uninitialized data bytes were printed.
2058 There were reports of junk text (parts of files, etc) appearing after
2061 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2063 PH/22 Added support for macro redefinition, and (re)definition in between
2064 driver and ACL definitions.
2066 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2067 forgetting to use the resulting value; it was using the unexpanded value.
2069 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2070 hadn't been configured. The fix is from Juergen Kreileder, who
2071 understands it better than I do:
2073 "Here's what I see happening with three configured cyrus_sasl
2074 authenticators configured (plain, login, cram-md5):
2076 On startup auth_cyrus_sasl_init() gets called for each of these.
2077 This means three calls to sasl_listmech() without a specified mech_list.
2078 => SASL tests which mechs of all available mechs actually work
2079 => three warnings about OTP not working
2080 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2082 With the patch, sasl_listmech() also gets called three times. But now
2083 SASL's mech_list option is set to the server_mech specified in the the
2084 authenticator. Or in other words, the answer from sasl_listmech()
2085 gets limited to just the mech you're testing for (which is different
2087 => the return list contains just 'plain' or 'login', 'cram-md5' or
2088 nothing depending on the value of ob->server_mech.
2090 I've just tested the patch: Authentication still works fine,
2091 unavailable mechs specified in the exim configuration are still
2092 caught, and the auth.log warnings about OTP are gone."
2094 PH/25 When debugging is enabled, the contents of the command line are added
2095 to the debugging output, even when log_selector=+arguments is not
2098 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2099 answer is "GNU", and only if the return is "GNU/something" is the answer
2102 PH/27 $acl_verify_message is now set immediately after the failure of a
2103 verification in an ACL, and so is available in subsequent modifiers. In
2104 particular, the message can be preserved by coding like this:
2106 warn !verify = sender
2107 set acl_m0 = $acl_verify_message
2109 Previously, $acl_verify_message was set only while expanding "message"
2110 and "log_message" when a very denied access.
2112 PH/28 Modified OS/os.c-Linux with
2114 -#ifndef OS_LOAD_AVERAGE
2115 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2117 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2118 nomenclature these days.)
2120 PH/29 Installed patch from the Sieve maintainer that adds the options
2121 sieve_useraddress and sieve_subaddress to the redirect router.
2123 PH/30 In these circumstances:
2124 . Two addresses routed to the same list of hosts;
2125 . First host does not offer TLS;
2126 . First host accepts first address;
2127 . First host gives temporary error to second address;
2128 . Second host offers TLS and a TLS session is established;
2129 . Second host accepts second address.
2130 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2131 and peerdn, if requested) that were in fact used only for the second
2134 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2135 attention to any local part prefix or suffix that was matched by the
2136 router that accepted the address. It now behaves in the same way as it
2137 does for delivery: the affixes are removed from the local part unless
2138 rcpt_include_affixes is set on the transport.
2140 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2141 timeout during the DATA phase of an incoming message.
2143 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2144 applied a patch sanctioned by the Sieve maintainer.
2146 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2147 the uid or gid is negative. A case of a negative gid caused this to be
2148 noticed. The fix allows for either to be negative.
2150 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2151 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2154 PH/36 The change PH/12 above was broken. Fixed it.
2156 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2157 the grounds that routing the same address twice would always produce the
2158 same answer. This might have been true once, but it is certainly no
2159 longer true now. Routing a child address may depend on the previous
2160 routing that produced that child. Some complicated redirection strategies
2161 went wrong when messages had multiple recipients, and made Exim's
2162 behaviour dependent on the order in which the addresses were given.
2164 I have moved the duplicate checking until after the routing is complete.
2165 Exim scans the addresses that are assigned to local and remote
2166 transports, and removes any duplicates. This means that more work will be
2167 done, as duplicates will always all be routed, but duplicates are
2168 presumably rare, so I don't expect this is of any significance.
2170 For deliveries to pipes, files, and autoreplies, the duplicate checking
2171 still happens during the routing process, since they are not going to be
2174 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2175 It corrects a timeout issue with spamd. This is Ian's comment: "The
2176 background is that sometimes spamd either never reads data from a
2177 connection it has accepted, or it never writes response data. The exiscan
2178 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2179 blindly assumes that writes won't block so it may never time out."
2181 PH/39 Allow G after quota size as well as K and M.
2183 PH/40 The value set for $authenticated_id in an authenticator may not contain
2184 binary zeroes or newlines because the value is written to log lines and
2185 to spool files. There was no check on this. Now the value is run through
2186 the string_printing() function so that such characters are converted to
2187 printable escape sequences.
2189 PH/41 $message_linecount is a new variable that contains the total number of
2190 lines in the message. Compare $body_linecount, which is the count for the
2193 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2194 bounce and delay warning messages, except in certain special cases, which
2197 (a) An SMTP error message from a remote host;
2198 (b) A message specified in a :fail: redirection;
2199 (c) A message specified in a "fail" command in a system filter;
2200 (d) A message specified in a FAIL return from the queryprogram router;
2201 (e) A message specified by the cannot_route_message router option.
2203 In these cases only, Exim does include the error details in bounce and
2204 warning messages. There are also a few cases where bland messages such
2205 as "unrouteable address" or "local delivery error" are given.
2207 PH/43 $value is now also set for the "else" part of a ${run expansion.
2209 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2210 being worked on, but at least Exim now implements the latest version to
2213 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2214 process to complete was treated as a delivery failure, a timeout while
2215 writing the message to the pipe was logged, but erroneously treated as a
2216 successful delivery. Such timeouts include transport filter timeouts. For
2217 consistency with the overall process timeout, these timeouts are now
2218 treated as errors, giving rise to delivery failures by default. However,
2219 there is now a new Boolean option for the pipe transport called
2220 timeout_defer, which, if set TRUE, converts the failures into defers for
2221 both kinds of timeout. A transport filter timeout is now identified in
2224 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2225 systems where "make" and "gmake" are different, calling "gmake" at top
2226 level broke things. I've arranged for the value of $(MAKE) to be passed
2227 from the Makefile to this script so that it can call the same version of
2231 A note about Exim versions 4.44 and 4.50
2232 ----------------------------------------
2234 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2235 changes of various kinds. As a consequence, a big documentation update was
2236 needed. This delayed the release for rather longer than seemed good, especially
2237 in the light of a couple of (minor) security issues. Therefore, the changes
2238 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2239 release. So 4.44 and 4.50 are in effect two different branches that both start
2242 I have left the 4.50 change log unchanged; it contains all the changes since
2243 4.43. The change log for 4.44 is below; many of its items are identical to
2244 those for 4.50. This seems to be the most sensible way to preserve the
2245 historical information.
2251 1. Minor wording change to the doc/README.SIEVE file.
2253 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2254 computation of the current number of files was incorrect.
2256 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2257 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2260 4. Give more explanation in the error message when the command for a transport
2261 filter fails to execute.
2263 5. There are several places where Exim runs a non-Exim command in a
2264 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2265 was being done only for the command run by the queryprogram router. It is
2266 now done for all such subprocesses. The other cases are: ${run, transport
2267 filters, and the commands run by the lmtp and pipe transports.
2269 6. Added CONFIGURE_GROUP build-time option.
2271 7. Some older OS have a limit of 256 on the maximum number of file
2272 descriptors. Exim was using setrlimit() to set 1000 as a large value
2273 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2274 systems. I've change it so that if it can't get 1000, it tries for 256.
2276 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2277 was an oversight, and furthermore, ever since the addition of extra
2278 controls (e.g. 4.43/32), the checks on when to allow different forms of
2279 "control" were broken. There should now be diagnostics for all cases when a
2280 control that does not make sense is encountered.
2282 9. Added the /retain_sender option to "control=submission".
2284 10. $recipients is now available in the predata ACL (oversight).
2286 11. Tidy the search cache before the fork to do a delivery from a message
2287 received from the command line. Otherwise the child will trigger a lookup
2288 failure and thereby defer the delivery if it tries to use (for example) a
2289 cached ldap connection that the parent has called unbind on.
2291 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2292 of $address_data from the recipient verification was clobbered by the
2293 sender verification.
2295 13. The value of address_data from a sender verification is now available in
2296 $sender_address_data in subsequent conditions in the ACL statement.
2298 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2300 15. Added a new option "connect=<time>" to callout options, to set a different
2303 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2304 was its contents. (It was OK if the option was not defined at all.)
2306 17. A "Completed" log line is now written for messages that are removed from
2307 the spool by the -Mrm option.
2309 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2310 information about exactly what failed.
2312 19. Added -dd to debug only the daemon process.
2314 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2315 handles timeouts, both on the server side and network timeouts. Renamed the
2316 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2318 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2319 It is now set to "smtps".
2321 22. $host_address is now set to the target address during the checking of
2322 ignore_target_hosts.
2324 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2325 being passed; this would have caused $sender_host_name to have been used if
2326 matching the list had actually called for a host name (not very likely,
2327 since this list is usually IP addresses). A host name is now passed as
2330 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2331 code that specifies a non-restarting handler (typically sigaction() in
2332 modern systems) in an attempt to fix a rare and obscure crash bug.
2334 25. Narrowed the window for a race in the daemon that could cause it to ignore
2335 SIGCHLD signals. This is not a major problem, because they are used only to
2336 wake it up if nothing else does.
2338 26. A malformed maildirsize file could cause Exim to calculate negative values
2339 for the mailbox size or file count. Odd effects could occur as a result.
2340 The maildirsize information is now recalculated if the size or filecount
2343 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2344 support for a long time. Removed HAVE_SYS_VFS_H.
2346 28. Installed the latest version of exipick from John Jetmore.
2348 29. In an address list, if the pattern was not a regular expression, an empty
2349 subject address (from a bounce message) matched only if the pattern was an
2350 empty string. Non-empty patterns were not even tested. This was the wrong
2351 because it is perfectly reasonable to use an empty address as part of a
2352 database query. An empty address is now tested by patterns that are
2353 lookups. However, all the other forms of pattern expect the subject to
2354 contain a local part and a domain, and therefore, for them, an empty
2355 address still always fails if the pattern is not itself empty.
2357 30. Exim went into a mad DNS loop when attempting to do a callout where the
2358 host was specified on an smtp transport, and looking it up yielded more
2359 than one IP address.
2361 31. Re-factored the code for checking spool and log partition space into a
2362 function that finds that data and another that does the check. The former
2363 is then used to implement four new variables: $spool_space, $log_space,
2364 $spool_inodes, and $log_inodes.
2366 32. The RFC2047 encoding function was originally intended for short strings
2367 such as real names; it was not keeping to the 75-character limit for
2368 encoded words that the RFC imposes. It now respects the limit, and
2369 generates multiple encoded words if necessary. To be on the safe side, I
2370 have increased the buffer size for the ${rfc2047: expansion operator from
2373 33. It is now permitted to omit both strings after an "if" condition; if the
2374 condition is true, the result is "true". As before, when the second string
2375 is omitted, a false condition yields an empty string. This makes it less
2376 cumbersome to write custom ACL and router conditions.
2378 34. Failure to deliver a bounce message always caused it to be frozen, even if
2379 there was an errors_to setting on the router. The errors_to setting is now
2382 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2383 canonical form (fully expanded) before being placed in
2384 $sender_host_address.
2386 36. The table in the code that translates DNS record types into text (T_A to
2387 "A" for instance) was missing entries for NS and CNAME. It is just possible
2388 that this could have caused confusion if both these types were looked up
2389 for the same domain, because the text type is used as part of Exim's
2390 per-process caching. But the chance of anyone hitting this buglet seems
2393 37. The dnsdb lookup has been extended in a number of ways.
2395 (1) There is a new type, "zns", which walks up the domain tree until it
2396 finds some nameserver records. It should be used with care.
2398 (2) There is a new type, "mxh", which is like "mx" except that it returns
2399 just the host names, not the priorities.
2401 (3) It is now possible to give a list of domains (or IP addresses) to be
2402 looked up. The behaviour when one of the lookups defers can be
2403 controlled by a keyword.
2405 (4) It is now possible to specify the separator character for use when
2406 multiple records are returned.
2408 38. The dnslists ACL condition has been extended: it is now possible to supply
2409 a list of IP addresses and/or domains to be looked up in a particular DNS
2412 39. Added log_selector=+queue_time_overall.
2414 40. When running the queue in the test harness, wait just a tad after forking a
2415 delivery process, to get repeatability of debugging output.
2417 41. Include certificate and key file names in error message when GnuTLS fails
2418 to set them up, because the GnuTLS error message doesn't include the name
2419 of the failing file when there is a problem reading it.
2421 42. Allow both -bf and -bF in the same test run.
2423 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2425 44. The "Exiscan patch" is now merged into the mainline Exim source.
2427 45. Sometimes the final signoff response after QUIT could fail to get
2428 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2429 < 0 before doing a fflush(). This bug looks as though it goes back to the
2430 introduction of TLS in release 3.20, but "sometimes" must have been rare
2431 because the tests only now provoked it.
2433 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2434 (this can affect the format of dates).
2436 47. exim_tidydb, when checking for the continued existence of a message for
2437 which it has found a message-specific retry record, was not finding
2438 messages that were in split spool directories. Consequently, it was
2439 deleting retry records that should have stayed in existence.
2441 48. Steve fixed some bugs in eximstats.
2443 49. The SPA authentication driver was not abandoning authentication and moving
2444 on to the next authenticator when an expansion was forced to fail,
2445 contradicting the general specification for all authenticators. Instead it
2446 was generating a temporary error. It now behaves as specified.
2448 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2449 (the order specifies the preference for clients). The order is now AES256,
2450 AES128, 3DES, ARCFOUR128.
2452 51. Small patch to Sieve code - explicitly set From: when generating an
2455 52. Exim crashed if a remote delivery caused a very long error message to be
2456 recorded - for instance if somebody sent an entire SpamAssassin report back
2457 as a large number of 550 error lines. This bug was coincidentally fixed by
2458 increasing the size of one of Exim's internal buffers (big_buffer) that
2459 happened as part of the Exiscan merge. However, to be on the safe side, I
2460 have made the code more robust (and fixed the comments that describe what
2463 53. Now that there can be additional text after "Completed" in log lines (if
2464 the queue_time_overall log selector is set), a one-byte patch to exigrep
2465 was needed to allow it to recognize "Completed" as not the last thing in
2468 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2469 patch that reportedly fixes this has been added. I am not expert enough to
2470 create a test for it. This is what the patch creator wrote:
2472 "I found a little strange behaviour of ldap code when working with
2473 Windows 2003 AD Domain, where users was placed in more than one
2474 Organization Units. When I tried to give exim partial DN, the exit code
2475 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2476 But simultaneously result of request was absolutely normal ldap result,
2477 so I produce this patch..."
2479 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2480 REFERENCE, so I have modified the code to exclude the patch when that macro
2483 55. Some experimental protocols are using DNS PTR records for new purposes. The
2484 keys for these records are domain names, not reversed IP addresses. The
2485 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2486 leaves it alone. Component reversal etc. now happens only for IP addresses.
2489 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2491 57. Double the size of the debug message buffer (to 2048) so that more of very
2492 long debug lines gets shown.
2494 58. The exicyclog utility now does better if the number of log files to keep
2495 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2497 59. Two changes related to the smtp_active_hostname option:
2499 (1) $smtp_active_hostname is now available as a variable.
2500 (2) The default for smtp_banner uses $smtp_active_hostname instead
2501 of $primary_hostname.
2503 60. The host_aton() function is supposed to be passed a string that is known
2504 to be a valid IP address. However, in the case of IPv6 addresses, it was
2505 not checking this. This is a hostage to fortune. Exim now panics and dies
2506 if the condition is not met. A case was found where this could be provoked
2507 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2508 components; fortuitously, this particular loophole had already been fixed
2509 by change 4.50/55 above.
2511 If there are any other similar loopholes, the new check in host_aton()
2512 itself should stop them being exploited. The report I received stated that
2513 data on the command line could provoke the exploit when Exim was running as
2514 exim, but did not say which command line option was involved. All I could
2515 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2516 running as the user.
2519 61. There was a buffer overflow vulnerability in the SPA authentication code
2520 (which came originally from the Samba project). I have added a test to the
2521 spa_base64_to_bits() function which I hope fixes it.
2524 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2525 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2527 63. The daemon start-up calls getloadavg() while still root for those OS that
2528 need the first call to be done as root, but it missed one case: when
2529 deliver_queue_load_max is set with deliver_drop_privilege. This is
2530 necessary for the benefit of the queue runner, because there is no re-exec
2531 when deliver_drop_privilege is set.
2533 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2534 This has been fixed.
2536 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2537 was in use, was not putting the data itself into the right store pool;
2538 consequently, it could be overwritten for a subsequent message in the same
2539 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2542 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2544 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2545 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2546 the function were treating the return as a boolean value, which happened to
2547 work because 0=false and not-0=true, but is not correct code.
2549 68. The host_aton() function was not handling scoped IPv6 addresses (those
2550 with, for example, "%eth0" on the end) correctly.
2552 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2553 negated items (that is, ~something) in unsigned ints. Some compilers
2554 apparently mutter when there is no cast.
2556 70. If an address verification called from an ACL failed, and did not produce a
2557 user-specific message (i.e. there was only a "system" message), nothing was
2558 put in $acl_verify_message. In this situation, it now puts the system
2561 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2562 session; change 4.31/43 added the unwanted input to the log line - except
2563 that it did not do this in the start of session case. It now does.
2565 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2566 This is wrong and can cause the other end to generate a synchronization
2567 error if it is another Exim or anything else that does the synchronization
2568 check. A QUIT command is no longer sent after a timeout.
2570 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2571 during host lookups.
2573 74. The defer_ok option of callout verification was not working if it was used
2574 when verifying addresses in header lines, that is, for this case:
2576 verify = header_sender/callout=defer_ok
2578 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2579 those file descriptors could be used for SMTP connections. If anything
2580 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2581 could be sent to the SMTP client, causing chaos. The daemon now opens
2582 stdin, stdout, and stderr to /dev/null when it puts itself into the
2585 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2586 log by default. The user can override this with suitable Perl magic.
2588 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2589 the log message when discard triggers, was not working for the DATA ACL or
2590 for the non-SMTP ACL.
2592 78. Error message wording change in sieve.c.
2594 79. If smtp_accept_max_per_host was set, the number of connections could be
2595 restricted to fewer than expected, because the daemon was trying to set up
2596 a new connection before checking whether the processes handling previous
2597 connections had finished. The check for completed processes is now done
2598 earlier. On busy systems, this bug wouldn't be noticed because something
2599 else would have woken the daemon, and it would have reaped the completed
2602 80. If a message was submitted locally by a user whose login name contained one
2603 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2604 It caused a spool format error. I have fixed the spool reading code. A
2605 related problem was that the "from" clause in the Received: line became
2606 illegal because of the space(s). It is now covered by ${quote_local_part.
2608 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2611 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2613 83. Patch from Sieve maintainer to fix unterminated string problem in
2614 "vacation" handling.
2616 84. Some minor changes to the Linux configuration files to help with other
2617 OS variants using glibc.
2619 85. One more patch for Sieve to update vacation handling to latest spec.
2622 ----------------------------------------------------
2623 See the note above about the 4.44 and 4.50 releases.
2624 ----------------------------------------------------
2630 1. Change 4.43/35 introduced a bug that caused file counts to be
2631 incorrectly computed when quota_filecount was set in an appendfile
2634 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2635 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2638 3. Give more explanation in the error message when the command for a transport
2639 filter fails to execute.
2641 4. There are several places where Exim runs a non-Exim command in a
2642 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2643 was being done only for the command run by the queryprogram router. It is
2644 now done for all such subprocesses. The other cases are: ${run, transport
2645 filters, and the commands run by the lmtp and pipe transports.
2647 5. Some older OS have a limit of 256 on the maximum number of file
2648 descriptors. Exim was using setrlimit() to set 1000 as a large value
2649 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2650 systems. I've change it so that if it can't get 1000, it tries for 256.
2652 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2653 was an oversight, and furthermore, ever since the addition of extra
2654 controls (e.g. 4.43/32), the checks on when to allow different forms of
2655 "control" were broken. There should now be diagnostics for all cases when a
2656 control that does not make sense is encountered.
2658 7. $recipients is now available in the predata ACL (oversight).
2660 8. Tidy the search cache before the fork to do a delivery from a message
2661 received from the command line. Otherwise the child will trigger a lookup
2662 failure and thereby defer the delivery if it tries to use (for example) a
2663 cached ldap connection that the parent has called unbind on.
2665 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2666 of $address_data from the recipient verification was clobbered by the
2667 sender verification.
2669 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2670 was its contents. (It was OK if the option was not defined at all.)
2672 11. A "Completed" log line is now written for messages that are removed from
2673 the spool by the -Mrm option.
2675 12. $host_address is now set to the target address during the checking of
2676 ignore_target_hosts.
2678 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2679 being passed; this would have caused $sender_host_name to have been used if
2680 matching the list had actually called for a host name (not very likely,
2681 since this list is usually IP addresses). A host name is now passed as
2684 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2685 code that specifies a non-restarting handler (typically sigaction() in
2686 modern systems) in an attempt to fix a rare and obscure crash bug.
2688 15. Narrowed the window for a race in the daemon that could cause it to ignore
2689 SIGCHLD signals. This is not a major problem, because they are used only to
2690 wake it up if nothing else does.
2692 16. A malformed maildirsize file could cause Exim to calculate negative values
2693 for the mailbox size or file count. Odd effects could occur as a result.
2694 The maildirsize information is now recalculated if the size or filecount
2697 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2698 support for a long time. Removed HAVE_SYS_VFS_H.
2700 18. Updated exipick to current release from John Jetmore.
2702 19. Allow an empty sender to be matched against a lookup in an address list.
2703 Previously the only cases considered were a regular expression, or an
2706 20. Exim went into a mad DNS lookup loop when doing a callout where the
2707 host was specified on the transport, if the DNS lookup yielded more than
2710 21. The RFC2047 encoding function was originally intended for short strings
2711 such as real names; it was not keeping to the 75-character limit for
2712 encoded words that the RFC imposes. It now respects the limit, and
2713 generates multiple encoded words if necessary. To be on the safe side, I
2714 have increased the buffer size for the ${rfc2047: expansion operator from
2717 22. Failure to deliver a bounce message always caused it to be frozen, even if
2718 there was an errors_to setting on the router. The errors_to setting is now
2721 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2722 canonical form (fully expanded) before being placed in
2723 $sender_host_address.
2725 24. Updated eximstats to version 1.33
2727 25. Include certificate and key file names in error message when GnuTLS fails
2728 to set them up, because the GnuTLS error message doesn't include the name
2729 of the failing file when there is a problem reading it.
2731 26. Expand error message when OpenSSL has problems setting up cert/key files.
2734 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2735 (this can affect the format of dates).
2737 28. exim_tidydb, when checking for the continued existence of a message for
2738 which it has found a message-specific retry record, was not finding
2739 messages that were in split spool directories. Consequently, it was
2740 deleting retry records that should have stayed in existence.
2742 29. eximstats updated to version 1.35
2743 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2744 1.35 - bugfix such that pie charts by volume are generated correctly
2746 30. The SPA authentication driver was not abandoning authentication and moving
2747 on to the next authenticator when an expansion was forced to fail,
2748 contradicting the general specification for all authenticators. Instead it
2749 was generating a temporary error. It now behaves as specified.
2751 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2752 (the order specifies the preference for clients). The order is now AES256,
2753 AES128, 3DES, ARCFOUR128.
2755 31. Small patch to Sieve code - explicitly set From: when generating an
2758 32. Exim crashed if a remote delivery caused a very long error message to be
2759 recorded - for instance if somebody sent an entire SpamAssassin report back
2760 as a large number of 550 error lines. This bug was coincidentally fixed by
2761 increasing the size of one of Exim's internal buffers (big_buffer) that
2762 happened as part of the Exiscan merge. However, to be on the safe side, I
2763 have made the code more robust (and fixed the comments that describe what
2766 33. Some experimental protocols are using DNS PTR records for new purposes. The
2767 keys for these records are domain names, not reversed IP addresses. The
2768 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2769 leaves it alone. Component reversal etc. now happens only for IP addresses.
2772 34. The host_aton() function is supposed to be passed a string that is known
2773 to be a valid IP address. However, in the case of IPv6 addresses, it was
2774 not checking this. This is a hostage to fortune. Exim now panics and dies
2775 if the condition is not met. A case was found where this could be provoked
2776 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2777 components; fortuitously, this particular loophole had already been fixed
2778 by change 4.50/55 or 4.44/33 above.
2780 If there are any other similar loopholes, the new check in host_aton()
2781 itself should stop them being exploited. The report I received stated that
2782 data on the command line could provoke the exploit when Exim was running as
2783 exim, but did not say which command line option was involved. All I could
2784 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2785 running as the user.
2788 35. There was a buffer overflow vulnerability in the SPA authentication code
2789 (which came originally from the Samba project). I have added a test to the
2790 spa_base64_to_bits() function which I hope fixes it.
2793 36. The daemon start-up calls getloadavg() while still root for those OS that
2794 need the first call to be done as root, but it missed one case: when
2795 deliver_queue_load_max is set with deliver_drop_privilege. This is
2796 necessary for the benefit of the queue runner, because there is no re-exec
2797 when deliver_drop_privilege is set.
2799 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2800 was in use, was not putting the data itself into the right store pool;
2801 consequently, it could be overwritten for a subsequent message in the same
2802 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2805 38. Sometimes the final signoff response after QUIT could fail to get
2806 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2807 < 0 before doing a fflush(). This bug looks as though it goes back to the
2808 introduction of TLS in release 3.20, but "sometimes" must have been rare
2809 because the tests only now provoked it.
2815 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2816 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2817 Now it returns an integer. A number of calls were still expecting a T/F
2818 return. Fortuitously, in all cases, the tests worked in OK situations,
2819 which is the norm. However, things would have gone wrong on any write
2820 failures on the smtp file descriptor. This function is used when sending
2821 messages over SMTP and also when doing verify callouts.
2823 2. When Exim is called to do synchronous delivery of a locally submitted
2824 message (the -odf or -odi options), it no longer closes stderr before doing
2827 3. Implemented the mua_wrapper option.
2829 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2831 5. Implemented the functions header_remove(), header_testname(),
2832 header_add_at_position(), and receive_remove_recipient(), and exported them
2835 6. If an ACL "warn" statement specified the addition of headers, Exim already
2836 inserted X-ACL-Warn: at the start if there was no header name. However, it
2837 was not making this test for the second and subsequent header lines if
2838 there were newlines in the string. This meant that an invalid header could
2839 be inserted if Exim was badly configured.
2841 7. Allow an ACL "warn" statement to add header lines at the start or after all
2842 the Received: headers, as well as at the end.
2844 8. Added the rcpt_4xx retry error code.
2846 9. Added postmaster_mailfrom=xxx to callout verification option.
2848 10. Added mailfrom=xxxx to the callout verification option, for verify=
2851 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2852 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2854 12. Inserted some casts to stop certain compilers warning when using pointer
2855 differences as field lengths or precisions in printf-type calls (mostly
2856 affecting debugging statements).
2858 13. Added optional readline() support for -be (dynamically loaded).
2860 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2861 same clock tick as a message's arrival, so that its received time was the
2862 same as the "first fail" time on the retry record, and that message
2863 remained on the queue past the ultimate address timeout, every queue runner
2864 would try a delivery (because it was past the ultimate address timeout) but
2865 after another failure, the ultimate address timeout, which should have then
2866 bounced the address, did not kick in. This was a "< instead of <=" error;
2867 in most cases the first failure would have been in the next clock tick
2868 after the received time, and all would be well.
2870 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2871 being recognized when the domain list was tested by the match_domain
2872 condition in an expansion string.
2874 16. Added the ${str2b64: operator.
2876 17. Exim was always calling setrlimit() to set a large limit for the number of
2877 processes, without checking whether the existing limit was already
2878 adequate. (It did check for the limit on file descriptors.) Furthermore,
2879 errors from getrlimit() and setrlimit() were being ignored. Now they are
2880 logged to the main and panic logs, but Exim does carry on, to try to do its
2881 job under whatever limits there are.
2883 18. Imported PCRE 5.0.
2885 19. Trivial typo in log message " temporarily refused connection" (the leading
2888 20. If the log selector return_path_on_delivery was set and an address was
2889 redirected to /dev/null, the delivery process crashed because it assumed
2890 that a return path would always be set for a "successful" delivery. In this
2891 case, the whole delivery is bypassed as an optimization, and therefore no
2894 21. Internal re-arrangement: the function for sending a challenge and reading
2895 a response while authentication was assuming a zero-terminated challenge
2896 string. It's now changed to take a pointer and a length, to allow for
2897 binary data in such strings.
2899 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2901 23. Exim was not respecting finduser_retries when seeking the login of the
2902 uid under which it was called; it was always trying 10 times. (The default
2903 setting of finduser_retries is zero.) Also, it was sleeping after the final
2904 failure, which is pointless.
2906 24. Implemented tls_on_connect_ports.
2908 25. Implemented acl_smtp_predata.
2910 26. If the domain in control=submission is set empty, Exim assumes that the
2911 authenticated id is a complete email address when it generates From: or
2912 Sender: header lines.
2914 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2915 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2916 chown and chgrp in /bin and hostname in /usr/bin.
2918 28. Exim was keeping the "process log" file open after each use, just as it
2919 does for the main log. This opens the possibility of it remaining open for
2920 long periods when the USR1 signal hits a daemon. Occasional processlog
2921 errors were reported, that could have been caused by this. Anyway, it seems
2922 much more sensible not to leave this file open at all, so that is what now
2925 29. The long-running daemon process does not normally write to the log once it
2926 has entered its main loop, and it closes the log before doing so. This is
2927 so that log files can straightforwardly be renamed and moved. However,
2928 there are a couple of unusual error situations where the daemon does write
2929 log entries, and I had neglected to close the log afterwards.
2931 30. The text of an SMTP error response that was received during a remote
2932 delivery was being truncated at 512 bytes. This is too short for some of
2933 the long messages that one sometimes sees. I've increased the limit to
2936 31. It is now possible to make retry rules that apply only when a message has a
2937 specific sender, in particular, an empty sender.
2939 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2940 it possible to be selective about when SMTP synchronization is enforced.
2942 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2944 32. Implemented hosts_connection_nolog.
2946 33. Added an ACL for QUIT.
2948 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2951 35. Added mailbox_size and mailbox_filecount to appendfile.
2953 36. Added control = no_multiline_responses to ACLs.
2955 37. There was a bug in the logic of the code that waits for the clock to tick
2956 in the case where the clock went backwards by a substantial amount such
2957 that the microsecond fraction of "now" was more than the microsecond
2958 fraction of "then" (but the whole seconds number was less).
2960 38. Added support for the libradius Radius client library this is found on
2961 FreeBSD (previously only the radiusclient library was supported).
2967 1. When certain lookups returned multiple values in the form name=value, the
2968 quoting of the values was not always being done properly. Specifically:
2969 (a) If the value started with a double quote, but contained no whitespace,
2971 (b) If the value contained whitespace other than a space character (i.e.
2972 tabs or newlines or carriage returns) it was not quoted.
2973 This fix has been applied to the mysql and pgsql lookups by writing a
2974 separate quoting function and calling it from the lookup code. The fix
2975 should probably also be applied to nisplus, ibase and oracle lookups, but
2976 since I cannot test any of those, I have not disturbed their existing code.
2978 2. A hit in the callout cache for a specific address caused a log line with no
2979 reason for rejecting RCPT. Now it says "Previous (cached) callout
2980 verification failure".
2982 3. There was an off-by-one bug in the queryprogram router. An over-long
2983 return line was truncated at 256 instead of 255 characters, thereby
2984 overflowing its buffer with the terminating zero. As well as fixing this, I
2985 have increased the buffer size to 1024 (and made a note to document this).
2987 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2988 when Exim is waiting for an SMTP response from a remote server, Exim
2989 restarts its select() call on the socket, thereby resetting its timeout.
2990 This is not a problem when such interrupts are rare. Somebody set up a cron
2991 job to run exiwhat every 2 minutes, which is less than the normal select()
2992 timeout (5 or 10 minutes). This meant that the select() timeout never
2993 kicked in because it was always reset. I have fixed this by comparing the
2994 time when an interrupt arrives with the time at the start of the first call
2995 to select(). If more time than the timeout has elapsed, the interrupt is
2996 treated as a timeout.
2998 5. Some internal re-factoring in preparation for the addition of Sieve
2999 extensions (by MH). In particular, the "personal" test is moved to a
3000 separate function, and given an option for scanning Cc: and Bcc: (which is
3001 not set for Exim filters).
3003 6. When Exim created an email address using the login of the caller as the
3004 local part (e.g. when creating a From: or Sender: header line), it was not
3005 quoting the local part when it contained special characters such as @.
3007 7. Installed new OpenBSD configuration files.
3009 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3010 try to make them clearer.
3012 9. Callout options, other than the timeout value, were being ignored when
3013 verifying sender addresses in header lines. For example, when using
3015 verify = header_sender/callout=no_cache
3017 the cache was (incorrectly) being used.
3019 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3020 only the Cygwin environment.
3022 11. When return_path_on_delivery was set as a log selector, if different remote
3023 addresses in the same message used different return paths and parallel
3024 remote delivery occurred, the wrong values would sometimes be logged.
3025 (Whenever a remote delivery process finished, the return path value from
3026 the most recently started remote delivery process was logged.)
3028 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3029 lines when AUTH and/or TLS are in use. This is the "received protocol"
3030 field. Exim used to use "asmtp" for authenticated SMTP, without any
3031 indication (in the protocol name) for TLS use. Now it follows the RFC and
3032 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3033 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3034 names appear in log lines as well as in Received: header lines.
3036 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3037 extensions, and comparison tests, and to fix some bugs.
3039 14. Changes to the "personal" filter test:
3041 (1) The test was buggy in that it was just doing the equivalent of
3042 "contains" tests on header lines. For example, if a user's address was
3043 anne@some.where, the "personal" test would incorrectly be true for
3045 To: susanne@some.where
3047 This test is now done by extracting each address from the header in turn,
3048 and checking the entire address. Other tests that are part of "personal"
3049 are now done using regular expressions (for example, to check local parts
3050 of addresses in From: header lines).
3052 (2) The list of non-personal local parts in From: addresses has been
3053 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3054 taken from the Sieve specification recommendations.
3056 (3) If the message contains any header line starting with "List-" it is
3057 treated as non-personal.
3059 (4) The test for "circular" in the Subject: header line has been removed
3060 because it now seems ill-conceived.
3062 15. Minor typos in src/EDITME comments corrected.
3064 16. Installed latest exipick from John Jetmore.
3066 17. If headers_add on a router specified a text string that was too long for
3067 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3068 of string_sprintf() is now avoided.
3070 18. $message_body_size was not set (it was always zero) when running the DATA
3071 ACL and the local_scan() function.
3073 19. For the "mail" command in an Exim filter, no default was being set for
3074 the once_repeat time, causing a random time value to be used if "once" was
3075 specified. (If the value happened to be <= 0, no repeat happened.) The
3076 default is now 0s, meaning "never repeat". The "vacation" command was OK
3077 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3078 (I found it when inspecting the code).
3080 20. There is now an overall timeout for performing a callout verification. It
3081 defaults to 4 times the callout timeout, which applies to individual SMTP
3082 commands during the callout. The overall timeout applies when there is more
3083 than one host that can be tried. The timeout is checked before trying the
3084 next host. This prevents very long delays if there are a large number of
3085 hosts and all are timing out (e.g. when the network connections are timing
3086 out). The value of the overall timeout can be changed by specifying an
3087 additional sub-option for "callout", called "maxwait". For example:
3089 verify = sender/callout=5s,maxwait=20s
3091 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3092 to the end before writing, but this should make it even safer).
3094 22. Exim was forgetting that it had advertised PIPELINING for the second and
3095 subsequent messages on an SMTP connection. It was also not resetting its
3096 memory on STARTTLS and an internal HELO.
3098 23. When Exim logs an SMTP synchronization error within a session, it now
3099 records whether PIPELINING has been advertised or not.
3101 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3102 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3103 rather than long int.
3105 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3107 26. Added the never_mail option to autoreply.
3113 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3114 crash if the getsockname() call failed; this can happen if a connection is
3115 closed very soon after it is established. The problem was simply in the
3116 order in which certain operations were done, causing Exim to try to write
3117 to the SMTP stream before it had set up the file descriptor. The bug has
3118 been fixed by making things happen in the correct order.
3124 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3125 before the connection was closed, thus losing the rejection response.
3127 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3128 some early Solaris releases, but causes trouble in current releases where
3129 socklen_t is defined.
3131 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3134 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3137 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3138 that releasing the top of it at the end releases what was used for sub-
3139 expansions (unless the block got too big). However, discard this block if
3140 the first thing is a variable or header, so that we can use its block when
3141 it is dynamic (useful for very large $message_headers, for example).
3143 6. Lookups now cache *every* query, not just the most recent. A new, separate
3144 store pool is used for this. It can be recovered when all lookup caches are
3145 flushed. Lookups now release memory at the end of their result strings.
3146 This has involved some general refactoring of the lookup sources.
3148 7. Some code has been added to the store_xxx() functions to reduce the amount
3149 of flapping under certain conditions.
3151 8. log_incoming_interface used to affect only the <= reception log lines. Now
3152 it causes the local interface and port to be added to several more SMTP log
3153 lines, for example "SMTP connection from", and rejection lines.
3155 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3157 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3159 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3160 could be overwritten at the end of the current message (or the start of a
3161 new message if it was set in a HELO ACL). The value is now preserved for
3162 the duration of the SMTP connection.
3164 12. If a transport had a headers_rewrite setting, and a matching header line
3165 contained an unqualified address, that address was qualified, even if it
3166 did not match any rewriting rules. The underlying bug was that the values
3167 of the flags that permit the existence of unqualified sender and recipient
3168 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3169 non-local messages, and by -bnq for local messages) were not being
3170 preserved with the message after it was received.
3172 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3173 "next input=" as part of the text comprising the host identity instead of
3174 the correct text. The code was using the same buffer for two different
3175 strings. However, depending on which order the printing function evaluated
3176 its arguments, the bug did not always show up. Under Linux, for example, my
3177 test suite worked just fine.
3179 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3180 doesn't work with some older versions of Perl. It has been changed to "my",
3181 which in any case is probably the better facility to use.
3183 15. A really picky compiler found some instances of statements for creating
3184 error messages that either had too many or two few arguments for the format
3187 16. The size of the buffer for calls to the DNS resolver has been increased
3188 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3189 for addresses that have a lot of PTR records. This alleviates a problem; it
3190 does not fully solve it.
3192 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3193 buffer now truncates the list and logs the incident, which is the same
3194 action as happens when Exim is looking up a host name and its aliases.
3195 Previously in this situation something unpredictable would happen;
3196 sometimes it was "internal error: store_reset failed".
3198 18. If a server dropped the connection unexpectedly when an Exim client was
3199 using GnuTLS and trying to read a response, the client delivery process
3200 crashed while trying to generate an error log message.
3202 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3203 string, for example:
3205 warn message = H1: something\nH2: something
3207 the text was added as a single header line from Exim's point of view
3208 though it ended up OK in the delivered message. However, searching for the
3209 second and subsequent header lines using $h_h2: did not work. This has been
3210 fixed. Similarly, if a system filter added multiple headers in this way,
3211 the routers could not see them.
3213 20. Expanded the error message when iplsearch is called with an invalid key to
3214 suggest using net-iplsearch in a host list.
3216 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3217 ACLs are no longer actually imposed (and a message to that effect is
3220 22. If a "gecos" field in a passwd entry contained escaped characters, in
3221 particular, if it contained a \" sequence, Exim got it wrong when building
3222 a From: or a Sender: header from that name. A second bug also caused
3223 incorrect handling when an unquoted " was present following a character
3224 that needed quoting.
3226 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3227 was not being matched caselessly.
3229 24. Arranged for all hyphens in the exim.8 source to be escaped with
3232 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3233 properly. Recipient callout cache records were still being keyed to include
3234 the sender, even when use_sender was set false. This led to far more
3235 callouts that were necessary. The sender is no longer included in the key
3236 when use_sender is false.
3238 26. Added "control = submission" modifier to ACLs.
3240 27. Added the ${base62d: operator to decode base 62 numbers.
3242 28. dnsdb lookups can now access SRV records.
3244 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3245 the configuration file.
3247 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3248 (-v) mode. This makes the output for a verbose queue run more intelligible.
3250 31. Added a use_postmaster feature to recipient callouts.
3252 32. Added the $body_zerocount variable, containing the number of binary zero
3253 bytes in the message body.
3255 33. The time of last modification of the "new" subdirectory is now used as the
3256 "mailbox time last read" when there is a quota error for a maildir
3259 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3261 35. Added +ignore_unknown as a special item in host lists.
3263 36. Code for decoding IPv6 addresses in host lists is now included, even if
3264 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3265 address was recognized as an IP address, but was then not correctly decoded
3266 into binary, causing unexpected and incorrect effects when compared with
3273 1. Very minor rewording of debugging text in manualroute to say "list of
3274 hosts" instead of "hostlist".
3276 2. If verify=header_syntax was set, and a header line with an unqualified
3277 address (no domain) and a large number of spaces between the end of the
3278 name and the colon was received, the reception process suffered a buffer
3279 overflow, and (when I tested it) crashed. This was caused by some obsolete
3280 code that should have been removed. The fix is to remove it!
3282 3. When running in the test harness, delay a bit after writing a bounce
3283 message to get a bit more predictability in the log output.
3285 4. Added a call to search_tidyup() just before forking a reception process. In
3286 theory, someone could use a lookup in the expansion of smtp_accept_max_
3287 per_host which, without the tidyup, could leave open a database connection.
3289 5. Added the variables $recipient_data and $sender_data which get set from a
3290 lookup success in an ACL "recipients" or "senders" condition, or a router
3291 "senders" option, similar to $domain_data and $local_part_data.
3293 6. Moved the writing of debug_print from before to after the "senders" test
3296 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3297 problems for message scanning, either using a data ACL, or using
3298 local_scan() because the Received: header was not generated till after they
3299 were called (in order to set the time as the time of reception completion).
3300 I have revised the way this works. The header is now generated after the
3301 body is received, but before the ACL or local_scan() are called. After they
3302 are run, the timestamp in the header is updated.
3308 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3309 before starting a queue runner without re-exec. This happened only when
3310 deliver_drop_privilege was set or when the Exim user was set to root. The
3311 effect of the bug was that timeouts during subsequent deliveries caused
3312 crashes instead of being properly handled. The handler is now left at its
3313 default (and expected) setting.
3315 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3316 message, again when deliver_drop_privilege is set or Exim is run as root.
3317 The bug described in (1) was not present in this case, but the tidying up
3318 of the other signals was missing. I have made the two cases consistent.
3320 3. The ignore_target_hosts setting on a manualroute router was being ignored
3321 for hosts that were looked up using the /MX notation.
3323 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3326 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3327 operated on the sender address. After changing the $sender_address to <>
3328 for the sender address verify, Exim was re-instated it as the original
3329 (before rewriting) address, but remembering that it had rewritten it, so it
3330 wasn't rewriting it again. This bug also had the effect of breaking the
3331 sender address verification caching when the sender address was rewritten.
3333 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3334 This has been changed so that if the ip literal address matches
3335 ignore_target_hosts, the router declines.
3337 7. Added expansion conditions match_domain, match_address, and match_local_
3338 part (NOT match_host).
3340 8. The placeholder for the Received: header didn't have a length field set.
3342 9. Added code to Exim itself and to exim_lock to test for a specific race
3343 condition that could lead to file corruption when using MBX delivery. The
3344 issue is with the lockfile that is created in /tmp. If this file is removed
3345 after a process has opened it but before that process has acquired a lock,
3346 there is the potential for a second process to recreate the file and also
3347 acquire a lock. This could lead to two Exim processes writing to the file
3348 at the same time. The added code performs the same test as UW imapd; it
3349 checks after acquiring the lock that its file descriptor still refers to
3350 the same named file.
3352 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3353 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3356 11. Added the smtp_active_hostname option. If used, this will typically be made
3357 to depend on the incoming interface address. Because $interface_address is
3358 not set up until the daemon has forked a reception process, error responses
3359 that can happen earlier (such as "too many connections") no longer contain
3362 12. If an expansion in a condition on a "warn" statement fails because a lookup
3363 defers, the "warn" statement is abandoned, and the next ACL statement is
3364 processed. Previously this caused the whole ACL to be aborted.
3366 13. Added the iplsearch lookup type.
3368 14. Added ident_timeout as a log selector.
3370 15. Added tls_certificate_verified as a log selector.
3372 16. Added a global option tls_require_ciphers (compare the smtp transport
3373 option of the same name). This controls incoming TLS connections.
3375 17. I finally figured out how to make tls_require_ciphers do a similar thing
3376 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3377 before starting the TLS session.
3379 18. Tabs are now shown as \t in -bP output.
3381 19. If the log selector return_path_on_delivery was set, Exim crashed when
3382 bouncing a message because it had too many Received: header lines.
3384 20. If two routers both had headers_remove settings, and the first one included
3385 a superfluous trailing colon, the final name in the first list and the
3386 first name in the second list were incorrectly joined into one item (with a
3387 colon in the middle).
3393 1. Added -C and -D options to the exinext utility, mainly to make it easier
3394 to include in the automated testing, but these could be helpful when
3395 multiple configurations are in use.
3397 2. The exinext utility was not formatting the output nicely when there was
3398 an alternate port involved in the retry record key, nor when there was a
3399 message id as well (for retries that were specific to a specific message
3400 and a specific host). It was also confused by IPv6 addresses, because of
3401 the additional colons they contain. I have fixed the IPv4 problem, and
3402 patched it up to do a reasonable job for IPv6.
3404 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3405 delivery, the log line now contains "pipelined" if PIPELINING was used.
3407 4. An SMTP transport process used to panic and die if the bind() call to set
3408 an explicit outgoing interface failed. This has been changed; it is now
3409 treated in the same way as a connect() failure.
3411 5. A reference to $sender_host_name in the part of a conditional expansion
3412 that was being skipped was still causing a DNS lookup. This no longer
3415 6. The def: expansion condition was not recognizing references to header lines
3416 that used bh_ and bheader_.
3418 7. Added the _cache feature to named lists.
3420 8. The code for checking quota_filecount in the appendfile transport was
3421 allowing one more file than it should have been.
3423 9. For compatibility with Sendmail, the command line option
3431 and sets the incoming protocol and host name (for trusted callers). The
3432 host name and its colon can be omitted when only the protocol is to be set.
3433 Note the Exim already has two private options, -pd and -ps, that refer to
3434 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3435 "s", but I don't think that's a major issue.
3437 10. A number of refactoring changes to the code, none of which should affect
3440 (a) The number of logging options was getting close to filling up the
3441 32-bit word that was used as a bit map. I have split them into two classes:
3442 those that are passed in the argument to log_write(), and those that are
3443 only ever tested independently outside of that function. These are now in
3444 separate 32-bit words, so there is plenty of room for expansion again.
3445 There is no change in the user interface or the logging behaviour.
3447 (b) When building, for example, log lines, the code previously used a
3448 macro that called string_cat() twice, in order to add two strings. This is
3449 not really sufficiently general. Furthermore, there was one instance where
3450 it was actually wrong because one of the argument was used twice, and in
3451 one call a function was used. (As it happened, calling the function twice
3452 did not affect the overall behaviour.) The macro has been replaced by a
3453 function that can join an arbitrary number of extra strings onto a growing
3456 (c) The code for expansion conditions now uses a table and a binary chop
3457 instead of a serial search (which was left over from when there were very
3458 few conditions). Also, it now recognizes conditions like "pam" even when
3459 the relevant support is not compiled in: a suitably worded error message is
3460 given if an attempt is made to use such a condition.
3462 11. Added ${time_interval:xxxxx}.
3464 12. A bug was causing one of the ddress fields not to be passed back correctly
3465 from remote delivery subprocesses. The field in question was not being
3466 subsequently used, so this caused to problems in practice.
3468 13. Added new log selectors queue_time and deliver_time.
3470 14. Might have fixed a bug in maildirsizefile handling that threw up
3471 "unexpected character" debug warnings, and recalculated the data
3472 unnecessarily. In any case, I expanded the warning message to give more
3475 15. Added the message "Restricted characters in address" to the statements in
3476 the default ACL that block characters like @ and % in local parts.
3478 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3479 Three changes have been made:
3481 (a) There was a serious bug; a negative response to MAIL caused the whole
3482 recipient domain to be cached as invalid, thereby blocking all messages
3483 to all local parts at the same domain, from all senders. This bug has
3484 been fixed. The domain is no longer cached after a negative response to
3485 MAIL if the sender used is not empty.
3487 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3490 (c) A new callout option, "use_sender" has been added for people who want
3491 the modified behaviour.
3497 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3500 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3501 indeed breaks things for older releases.
3503 3. Added additional logging to the case where there is a problem reading data
3504 from a filter that is running in a subprocess using a pipe, in order to
3505 try to track down a specific problem.
3507 4. Testing facility fudge: when running in the test harness and attempting
3508 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3509 getting "No route to host". Convert this to a timeout.
3511 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3514 6. Some OS don't have socklen_t but use size_t instead. This affects the
3515 fifth argument of getsockopt() amongst other things. This is now
3516 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3517 can be set for individual OS. I have set it for SunOS5, OSF1, and
3518 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3519 some earlier ones do not.
3521 7. Change 4.30/15 was not doing the test caselessly.
3523 8. The standard form for an IPv6 address literal was being rejected by address
3524 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3525 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3526 this, as well as the form without the "IPv6" on the front (but only when
3527 address literals are enabled, of course).
3529 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3531 10. Exim crashed if a message with an empty sender address specified by -f
3532 encountered a router with an errors_to setting. This could be provoked only
3533 by a command such as
3537 where an empty string was supplied; "<>" did not hit this bug.
3539 11. Installed PCRE release 4.5.
3541 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3542 remained set. It is now erased.
3544 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3545 times from message ids (which are base 36 rather than the normal 62).
3547 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3548 were being counted as actual protocol errors, and logged if the log
3549 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3550 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3551 and DATA following a set of rejected RCPTs do not count as protocol errors.
3552 In other words, Exim assumes they were pipelined, though this may not
3553 actually be the case. Of course, in all cases the client gets an
3554 appropriate error code.
3556 15. If a lookup fails in an ACL condition, a message about the failure may
3557 be available; it is used if testing the ACL cannot continue, because most
3558 such messages specify what the cause of the deferral is. However, some
3559 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3560 that caused an old message to be retained and used if a later statement
3561 caused a defer, replacing the real cause of the deferral.
3563 16. If an IP address had so many PTR records that the DNS lookup buffer
3564 was not large enough to hold them, Exim could crash while trying to process
3565 the truncated data. It now detects and logs this case.
3567 17. Further to 4.21/58, another change has been made: if (and only if) the
3568 first line of a message (the first header line) ends with CRLF, a bare LF
3569 in a subsequent header line has a space inserted after it, so as not to
3570 terminate the header.
3572 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3573 unnecessarily, used atoi() instead of strtol(), and didn't check the
3574 termination when getting file sizes from file names by regex.
3576 19. Completely re-implemented the support for maildirsize files, in the light
3577 of a number of problems with the previous contributed implementation
3578 (4.30/29). In particular:
3580 . If the quota is zero, the maildirsize file is maintained, but no quota is
3583 . If the maildir directory does not exist, it is created before any attempt
3584 to write a maildirsize file.
3586 . The quota value in the file is just a cache; if the quota is changed in
3587 the transport, the new value overrides.
3589 . A regular expression is available for excluding directories from the
3592 20. The autoreply transport checks the characters in options that define the
3593 message's headers; it allows continued headers, but it was checking with
3594 isspace() after an embedded newline instead of explicitly looking for a
3597 21. If all the "regular" hosts to which an address was routed had passed their
3598 expiry times, and had not reached their retry times, the address was
3599 bounced, even if fallback hosts were defined. Now Exim should go on to try
3602 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3603 equivalent code in the SMTP transport. Some hosts send humungous responses
3604 to HELO/EHLO, more than 1024 it seems.
3606 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3607 gives compiler warnings in some environments. I've now done it "properly",
3610 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3611 (because of problems with the built-in one) was declared to return uschar *
3612 instead of char *, causing compiler failure.
3614 25. Fixed a file descriptor leak when processing alias/forward files.
3616 26. Fixed a minor format string issue in dbfn.c.
3618 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3620 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3621 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3624 29. When a sender address is verified, it is cached, to save repeating the test
3625 when there is more than one recipient in a message. However, when the
3626 verification involves a callout, it is possible for different callout
3627 options to be set for different recipients. It is too complicated to keep
3628 track of this in the cache, so now Exim always runs a verification when a
3629 callout is required, relying on the callout cache for the optimization.
3630 The overhead is duplication of the address routing, but this should not be
3633 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3634 to be verified with callout=postmaster, and the main callout worked but the
3635 postmaster check failed, the verification correctly failed. However, if a
3636 subsequent RCPT command asked for sender verification *without* the
3637 postmaster check, incorrect caching caused this verification also to fail,
3640 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3641 it was not caching the DNS options (qualify_single, search_parents) that
3642 were used when the lookup failed. A subsequent lookup with different
3643 options therefore always gave the same answer, though there were cases
3644 where it should not have. (Example: a "domains = !$mx_any" option on a
3645 dnslookup router: the "domains" option is always processed without any
3646 widening, but the router might have qualify_single set.) Now Exim uses the
3647 cached value only when the same options are set.
3649 32. Added John Jetmore's "exipick" utility to the distribution.
3651 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3652 than a timeout (e.g. a certificate is required, and is not provided), an
3653 Exim server now closes the connection immediately. Previously it waited for
3654 the client to close - but if the client is SSL, it seems that they each
3655 wait for each other, leading to a delay before one of them times out.
3657 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3658 maintained 0.8.x compatibility because I don't think many are using it, and
3659 it is clearly obsolete.
3661 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3664 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3665 client certificate was expired. A simple patch fixes this, though I don't
3666 understand the full logic of why the verify callback is called multiple
3669 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3670 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3671 which causes problems with some clients (such as the Certicom SSL Plus
3672 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3673 disables the coutermeasure allowing Eudora to connect."
3675 38. Exim was not checking that a write() to a log file succeeded. This could
3676 lead to Bad Things if a log got too big, in particular if it hit a file
3677 size limit. Exim now panics and dies if it cannot write to a log file, just
3678 as it does if it cannot open a log file.
3680 39. Modified OS/Makefile-Linux so that it now contains
3682 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3684 The two -D definitions ensure that Exim is compiled with large file
3685 support, which makes it possible to handle log files that are bigger than
3688 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3689 instance) a domain was checked against a named list that involved a lookup,
3690 causing $domain_data to be set, then another domain was checked against the
3691 same list, then the first domain was re-checked, the value of $domain_data
3692 after the final check could be wrong. In particular, if the second check
3693 failed, it could be set empty. This bug probably also applied to
3696 41. The strip_trailing_dot option was not being applied to the address given
3697 with the -f command-line option.
3699 42. The code for reading a message's header from the spool was incrementing
3700 $received_count, but never initializing it. This meant that the value was
3701 incorrect (doubled) while delivering a message in the same process in which
3702 it was received. In the most common configuration of Exim, this never
3703 happens - a fresh exec is done - but it can happen when
3704 deliver_drop_privilege is set.
3706 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3707 it now includes up to 150 characters of the unexpected data in the log
3710 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3711 and building data strings. The size of both of these buffers was 10 000
3712 bytes - far larger than anybody would *ever* want, thought I. Needless to
3713 say, somebody hit the limit. I have increased the maximum line length to
3714 20 000 and the maximum data length of concatenated lines to 100 000. I have
3715 also fixed two bugs, because there was no checking on these buffers. Tsk,
3716 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3717 buffer is too small.
3719 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3720 lsearch lookups. Now it does.
3722 46. When parsing a route_list item in a manualroute router, a fixed-length
3723 buffer was used for the list of hosts. I made this 1024 bytes long,
3724 thinking that nobody would ever have a list of hosts that long. Wrong.
3725 Somebody had a whole pile of complicated expansion conditions, and the
3726 string was silently truncated, leading to an expansion error. It turns out
3727 that it is easier to change to an unlimited length (owing to other changes
3728 that have happened since this code was originally written) than to build
3729 structure for giving a limitation error. The length of the item that
3730 expands into the list of hosts is now unlimited.
3732 47. The lsearch lookup could not handle data where the length of text line was
3733 more than 4095 characters. Such lines were truncated, leading to shortened
3734 data being returned. It should now handle lines of any length.
3736 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3737 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3740 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3741 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3742 no longer happens in comment lines. A list of replacements is now placed
3743 at the head of all of the source files, except those whose only change is
3744 to replace PERL_COMMAND in the very first #! line.
3746 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3747 messages on the queue, with a bottom-up merge sort, using code contributed
3748 by Michael Haardt. This should make operations like -bp somewhat faster on
3749 large queues. It won't affect queue runners, except when queue_run_in_order
3752 51. Installed eximstats 1.31 in the distribution.
3754 52. Added support for SRV lookups to the dnslookup router.
3756 53. If an ACL referred to $message_body or $message_body_end, the value was not
3757 reset for any messages that followed in the same SMTP session.
3759 54. The store-handling optimization for building very long strings was not
3760 differentiating between the different store pools. I don't think this
3761 actually made any difference in practice, but I've tidied it.
3763 55. While running the routers to verify a sender address, $sender_address
3764 was still set to the sender address. This is wrong, because when routing to
3765 send a bounce to the sender, it would be empty. Therefore, I have changed
3766 it so that, while verifying a sender address, $sender_address is set to <>.
3767 (There is no change to what happens when verifying a recipient address.)
3769 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3770 target A or AAAA records (if not already returned) without resetting the
3771 qualify_single or search_parents options of the DNS resolver. These are
3772 inappropriate in this case because the targets of MX and SRV records must
3773 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3774 target that, when qualified, matched something in the local domain. These
3775 two options are now turned off when doing these lookups.
3777 57. It seems that at least some releases of Reiserfs (which does not have the
3778 concept of a fixed number of inodes) returns zero and not -1 for the
3779 number of available inodes. This interacted badly with check_spool_inodes,
3780 which assumed that -1 was the "no such thing" setting. What I have done is
3781 to check that the total number of inodes is greater than zero before doing
3782 the test of how many are available.
3784 58. When a "warn" ACL statement has a log_message modifier, the message is
3785 remembered, and not repeated. This is to avoid a lot of repetition when a
3786 message has many recipients that cause the same warning to be written.
3787 Howewer, Exim was preserving the list of already written lines for an
3788 entire SMTP session, which doesn't seem right. The memory is now reset if a
3789 new message is started.
3791 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3792 debugging output unless log_rewrite was also set.
3794 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3795 of (int)(handle) when we know that handle contains (void *)(-1).
3797 61. The Exim daemon panic-logs an error return when it closes the incoming
3798 connection. However "connection reset by peer" seems to be common, and
3799 isn't really an error worthy of noting specially, so that particular error
3802 62. When Exim is trying to find all the local interfaces, it used to panic and
3803 die if the ioctl to get the interface flags failed. However, it seems that
3804 on at least one OS (Solaris 9) it is possible to have an interface that is
3805 included in the list of interfaces, but for which you get a failure error
3806 for this call. This happens when the interface is not "plumbed" into a
3807 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3808 failure of the "get flags" call assumes that the interface is down.
3810 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3811 makes life easier for people who are doing arithmetic on fields extracted
3812 from dates, where you often get leading zeros that should not be
3813 interpreted as octal.
3815 64. Added qualify_domain to the redirect router, to override the global
3818 65. If a pathologically long header line contained very many addresses (the
3819 report of this problem mentioned 10 000) and each of them was rewritten,
3820 Exim could use up a very large amount of memory. (It kept on making new
3821 copies of the header line as it rewrote, and never released the old ones.)
3822 At the expense of a bit more processing, the header rewriting function has
3823 been changed so that it no longer eats memory in this way.
3825 66. The generation of the Received: header has been moved from the time that a
3826 message starts to be received, to the time that it finishes. The timestamp
3827 in the Received: header should now be very close to that of the <= log
3828 line. There are two side-effects of this change:
3830 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3831 logged header lines no longer include the local Received: line, because
3832 it has not yet been created. The same applies to a copy of the message
3833 that is returned to a non-SMTP sender when a message is rejected.
3835 (b) When a filter file is tested using -bf, no additional Received: header
3836 is added to the test message. After some thought, I decided that this
3839 This change does not affect the value of $received_for. It is still set
3840 after address rewriting, but before local_scan() is called.
3842 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3844 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3845 gave an unhelpful panic error message, and a defer error. I have managed to
3846 change this behaviour so that it now rejects any supplied certificate,
3847 which seems right, as the list of acceptable certificates is empty.
3849 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3850 gave an unhelpful defer error. I have not managed to make this reject any
3851 supplied certificates, but the error message it gives is "no certificate
3852 supplied", which is not helpful.
3854 70. exigrep's output now also includes lines that are not associated with any
3855 message, but which match the given pattern. Implemented by a patch from
3856 Martin Sluka, which also tidied up the Perl a bit.
3858 71. Recipient callout verification, like sender verification, was using <> in
3859 the MAIL FROM command. This isn't really the right thing, since the actual
3860 sender may affect whether the remote host accepts the recipient or not. I
3861 have changed it to use the actual sender in the callout; this means that
3862 the cache record is now keyed on a recipient/sender pair, not just the
3863 recipient address. There doesn't seem to be a real danger of callout loops,
3864 since a callout by the remote host to check the sender would use <>.
3865 [SEE ABOVE: changed after hitting problems.]
3867 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3868 temporary errors. However, in the case of such a code being given after
3869 the end of a data transmission (i.e. after ".") Exim was failing to write
3870 a retry record for the message. (Yes, there was some broken host that was
3871 actually sending 8xx at this point.)
3873 73. An unknown lookup type in a host list could cause Exim to panic-die when
3874 the list was checked. (An example that provoked this was putting <; in the
3875 middle of a list instead of at the start.) If this happened during a DATA
3876 ACL check, a -D file could be left lying around. This kind of configuration
3877 error no longer causes Exim to die; instead it causes a defer errror. The
3878 incident is still logged to the main and panic logs.
3880 74. Buglet left over from Exim 3 conversion. The message "too many messages
3881 in one connection" was written to the rejectlog but not the mainlog, except
3882 when address rewriting (yes!) was being logged.
3884 75. Added write_rejectlog option.
3886 76. When a system filter was run not as root (that is, when system_filter_user
3887 was set), the values of the $n variables were not being returned to the
3888 main process; thus, they were not subsequently available in the $sn
3891 77. Added +return_path_on_delivery log selector.
3893 78. A connection timeout was being treated differently from recipients deferred
3894 when testing hosts_max_try with a message that was older than the host's
3895 retry timeout. (The host should not be counted, thus allowing all hosts to
3896 be tried at least once before bouncing.) This may have been the cause of an
3897 occasionally reported bug whereby a message would remain on the queue
3898 longer than the retry timeout, but would be bounced if a delivery was
3899 forced. I say "may" because I never totally pinned down the problem;
3900 setting up timeout/retry tests is difficult. See also the next item.
3902 79. The ultimate address timeout was not being applied to errors that involved
3903 a combination of host plus message (for example, a timeout on a MAIL
3904 command). When an address resolved to a number of possible hosts, and they
3905 were not all tried for each delivery (e.g. because of hosts_max_try), a
3906 message could remain on the queue longer than the retry timeout.
3908 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3911 81. Fixed an obscure SMTP outgoing bug which required at least the following
3912 conditions: (a) there was another message waiting for the same server;
3913 (b) the server returned 5xx to all RCPT commands in the first message so
3914 that the message was not completed; (c) the server dropped the connection
3915 or gave a negative response to the RSET that Exim sends to abort the
3916 transaction. The observed case was a dropped connection after DATA that had
3917 been sent in pipelining mode. That is, the server had advertised PIPELINING
3918 but was not implementing it correctly. The effect of the bug was incorrect
3919 behaviour, such as trying another host, and this could lead to a crash.
3925 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3926 and daemon.c were passed as pointers to ints; they should have been
3927 pointers to socklen_t variables (which are typically unsigned ints).
3929 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3932 3. Fixed a really odd bug that affected only the testing scheme; patching a
3933 certain fixed string in the binary changed the value of another string that
3934 happened to be identical to the end of the original first string.
3936 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3937 name", it returns that address as the IP address. On some operating
3938 systems (e.g. Solaris), it also passes back the IP address string as the
3939 "host name". However, on others (e.g. Linux), it passes back an empty
3940 string. Exim wasn't checking for this, and was changing the host name to an
3941 empty string, assuming it had been canonicized.
3943 5. Although rare, it is permitted to have more than one PTR record for a given
3944 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3945 all the names associated with an address, because they do in Solaris.
3946 However, it seems that they do not in Linux for data that comes from the
3947 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3948 I found this out when I moved to a new Linux workstation and tried to run
3949 the Exim test suite.
3951 To get round this problem I have changed the code so that it now does its
3952 own call to the DNS to look up PTR records when searching for a host name.
3953 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3954 addresses that are only in /etc/hosts are still found.
3956 This behaviour is, however, controlled by an option called host_lookup_
3957 order, which defaults to "bydns:byaddr". If people want to use the other
3958 order, or indeed, just use one or the other means of lookup, they can
3959 specify it in this variable.
3961 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3962 some operating systems, this comes back from gethostbyaddr() as an empty
3963 string, and this is what Exim used to test for. However, it seems that in
3964 other systems, "." is yielded. Exim now tests for this case too.
3966 7. The values of check_spool_space and check_log_space are now held internally
3967 as a number of kilobytes instead of an absolute number of bytes. If a
3968 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3969 kilobyte. This means that much larger values can be stored.
3971 8. Exim monitor: an attempt to get the action menu when not actually pointing
3972 at a message produces an empty menu entitled "No message selected". This
3973 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3974 no entries in it ("Shell widget menu has zero width and/or height"). So I
3975 have added a single, blank menu entry in this case.
3977 9. Added ${quote_local_part.
3979 10. MIME decoding is now applied to the contents of Subject: header lines when
3982 11. Now that a reference to $sender_host_address automatically causes a reverse
3983 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3984 host lookup before query-style lookups in lists that might use this
3985 variable. This has therefore been abolished, and the "net-" prefix is no
3986 longer necessary for query-style lookups.
3988 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3989 have been a typo for LFLAGS, so it has been changed.
3991 13. The install script calls Exim with "-C /dev/null" in order to find the
3992 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3993 to be output. Howeve, since Exim outputs its version number before the
3994 error, it didn't break the script. It just looked ugly. I fixed this by
3995 always allowing "-C /dev/null" if the caller is root.
3997 14. Ignore overlarge ACL variable number when reading spool file - insurance
3998 against a later release with more variables having written the file.
4000 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4001 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4002 this, as well as the form without the "IPv6" on the front.
4004 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4005 OS/Makefile-Darwin file.
4007 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4008 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4010 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4011 variable. However, some OS use ints for inodes. I've added cast to long int
4012 to get rid of the compiler warning.
4014 19. I had forgotten to lock out "/../" in configuration file names when
4015 ALT_CONFIG_PREFIX was set.
4017 20. Routers used for verification do not need to specify transports. However,
4018 if such a router generated a host list, and callout was configured, Exim
4019 crashed, because it could not find a port number from the (non-existent)
4020 transport. It now assumes port 25 in this circumstance.
4022 21. Added the -t option to exigrep.
4024 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4025 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4026 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4027 them is set, LOOKUP_LSEARCH is forced.
4029 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4031 24. Added sender and host information to the "rejected by local_scan()" log
4032 line; previously there was no indication of these.
4034 25. Added .include_if_exists.
4036 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4037 Linux. It turns out that not all file systems support this. Apparently some
4038 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4039 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4040 not supported on the file descriptor, is now ignored when Exim is trying to
4041 sync a directory. This applies only to Linux.
4043 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4045 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4046 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4049 29. Added support for maildirsize files from supplied patch (modified a bit).
4051 30. The use of :fail: followed by an empty string could lead Exim to respond to
4052 sender verification failures with (e.g.):
4054 550 Verification failed for <xxx>
4055 550 Sender verify failed
4057 where the first response line was missing the '-' that indicates it is not
4058 the final line of the response.
4060 31. The loop for finding the name of the user that called Exim had a hardwired
4061 limit of 10; it now uses the value of finduser_retries, which is used for
4062 all other user lookups.
4064 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4067 33. Exim was neglecting to zero errno before one call of strtol() when
4068 expanding a string and expecting an integer value. On some systems this
4069 resulted in spurious "integer overflow" errors. Also, it was casting the
4070 result into an int without checking.
4072 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4073 did not work. The code looks as if it has *never* worked, though it appears
4074 to have been documented since at least releast 1.62. I have made it work.
4076 35. The "timeout_DNS" error in retry rules, also documented since at least
4077 1.62, also never worked. As it isn't clear exactly what this means, and
4078 clearly it isn't a major issue, I have abolished the feature by treating it
4079 as "timeout", and writing a warning to the main and panic logs.
4081 36. The display of retry rules for -brt wasn't always showing the error code
4084 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4085 timeout_connect_A, timeout_connect_MX.
4087 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4088 to the empty sender.
4090 39. The daemon was not analysing the content of -oX till after it had closed
4091 stderr and disconnected from the controlling terminal. This meant that any
4092 syntax errors were only noted on the panic log, and the return code from
4093 the command was 0. By re-arranging the code a little, I've made the
4094 decoding happen first, so such errors now appear on stderr, and the return
4095 code is 1. However, the actual setting up of the sockets still happens in
4096 the disconnected process, so errors there are still only recorded on the
4099 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4100 connections (as happens on some IP stacks) was logged at start up time as
4101 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4102 it from "IPv6 and IPv4", which means that two separate sockets are being
4105 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4106 says whether AF_INET or AF_INET6 was passed as an argument.
4108 42. Exiwhat output was messed up when time zones were included in log
4111 43. Exiwhat now gives more information about the daemon's listening ports,
4112 and whether -tls-on-connect was used.
4114 44. The "port" option of the smtp transport is now expanded.
4116 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4117 silently ignored. Now an error message is written to the main and panic
4120 46. There's a new ACL modifier called "logwrite" which writes to a log file
4121 as soon as it is encountered.
4123 47. Added $local_user_uid and $local_user_gid at routing time.
4125 48. Exim crashed when trying to verify a sender address that was being
4128 49. Exim was recognizing only a space character after ".include". It now also
4129 recognizes a tab character.
4131 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4132 extracting the relevant information from the specification. The man page no
4133 longer contains scrambled data for the -d option, and I've added a section
4134 at the front about calling Exim under different names.
4136 51. Added "extra_headers" argument to the "mail" command in filter files.
4138 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4141 53. Installed eximstats 1.29.
4143 54. Added transport_filter_timeout as a generic transport option.
4145 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4146 Cc: header lines. This was required by RFC 822, but it not required by RFC
4149 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4150 incoming messages that did not have them. Now it does so only if the
4151 message originates locally, that is, if there is no associated remote host
4152 address. When Resent- header lines are present, this applies to the Resent-
4153 lines rather than the non-Resent- lines.
4155 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4156 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4158 58. Messages for configuration errors now include the name of the main
4159 configuration file - useful now that there may be more than one file in a
4160 list (.included file names were always shown).
4162 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4163 for those rare installations that do not start the daemon as root or run it
4164 setuid root. I've cut out the call to initgroups() if the daemon is not
4167 60. The Exim user and group can now be bound into the binary as text strings
4168 that are looked up at the start of Exim's processing.
4170 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4172 62. Added $mailstore_basename variable.
4174 63. Installed patch to sieve.c from Michael Haardt.
4176 64. When Exim failed to open the panic log after failing to open the main log,
4177 the original message it was trying to log was written to stderr and debug
4178 output, but if they were not available (the usual case in production), it
4179 was lost. Now it is written to syslog before the two lines that record the
4180 failures to open the logs.
4182 65. Users' Exim filters run in subprocesses under the user's uid. It is
4183 possible for a "deliver" command or an alias in a "personal" command to
4184 provoke an address rewrite. If logging of address rewriting is configured,
4185 this fails because the process is not running as root or exim. There may be
4186 a better way of dealing with this, but for the moment (because 4.30 needs
4187 to be released), I have disabled address rewrite logging when running a
4188 filter in a non-root, non-exim process.
4194 1. The buildconfig auxiliary program wasn't quoting the value set for
4195 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4196 not defined. This bug was masked in 4.22 by the effect that was fixed in
4199 2. Some messages that were rejected after a message id was allocated were
4200 shown as "incomplete" by exigrep. It no longer does this for messages that
4201 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4203 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4204 have allow_domain_literals set, the ID did not get logged in the <= line.
4205 Domain literals are now always recognized in Message-ID: header lines.
4207 4. The first argument for a ${extract expansion item is the key name or field
4208 number. Leading and trailing spaces in this item were not being ignored,
4209 causing some misleading effects.
4211 5. When deliver_drop_privilege was set, single queue runner processes started
4212 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4213 same command in the process it spins off) were not dropping privilege.
4215 6. When the daemon running as "exim" started a queue runner, it always
4216 re-executed Exim in the spun-off process. This is a waste of effort when
4217 deliver_drop_privilege is set. The new process now just calls the
4218 queue-runner function directly.
4224 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4227 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4228 ignored. Though the use of -oP was forcing the writing of a pid file, it
4229 was always written to the default place.
4231 3. If the message "no IP address found for host xxxx" is generated during
4232 incoming verification, it is now followed by identification of the incoming
4233 connection (so you can more easily find what provoked it).
4235 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4237 5. Added some features to "harden" Exim a bit more against certain attacks:
4239 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4240 be put in Local/Makefile. This is like the never_users runtime option,
4241 but it cannot be overridden. The default setting is "root".
4243 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4244 prefix string with which any file named in a -C command line option
4247 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4248 is retained for -C and -D only if the caller of Exim is root. Without
4249 it, the exim user may also use -C and -D and retain privilege.
4251 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4252 command line option is disabled.
4254 6. Macro names set by the -D option must start with an upper case letter, just
4255 like macro names defined in the configuration file.
4257 7. Added "dereference=" facility to LDAP.
4259 8. Two instances of the typo "uknown" in the source files are fixed.
4261 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4262 the Configure-Makefile script screwed up while processing it.
4264 10. Incorporated PCRE 4.4.
4266 11. The SMTP synchronization check was not operating right at the start of an
4267 SMTP session. For example, it could not catch a HELO sent before the client
4268 waited for the greeting. There is now a check for outstanding input at the
4269 point when the greeting is written. Because of the duplex, asynchronous
4270 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4271 way, but not yet received, when the check is performed.
4273 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4274 on TCP/IP sockets, because this apparently causes some broken clients to
4277 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4278 unchanged) from the Cygwin maintainer.
4280 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4281 was supported instead of testing for mbx. Effectively a typo.
4283 15. The spa authenticator server code was not checking that the input it
4284 received was valid base64.
4286 16. The debug output line for the "set" modifier in ACLs was not showing the
4287 name of the variable that was being set.
4289 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4291 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4292 lookup on its own. Something else was needed to trigger the lookup. For
4293 example, a match in host_lookup or the need for a host name in a host list.
4294 Now, if $sender_host_name is referenced and the host name has not yet been
4295 looked up, a lookup is performed. If the lookup fails, the variable remains
4296 empty, and $host_lookup_failed is set to "1".
4298 19. Added "eqi" as a case-independent comparison operator.
4300 20. The saslauthd authentication condition could segfault if neither service
4301 nor realm was specified.
4303 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4304 error message that was logged was misleading, and incoming SMTP
4305 connections were dropped. The message is now more accurate, and temporary
4306 errors are given to SMTP connections.
4308 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4309 (except RSET) with a 421 error, until QUIT is received. However, it was
4310 failing to send a response to QUIT.
4312 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4313 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4314 ACL is now run after the helo_try_verify_hosts code.
4316 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4317 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4318 case-independent, so other case variants are also recognized). Apparently
4319 some systems use these upper case variants.
4321 25. If more than two messages were waiting for the same host, and a transport
4322 filter was specified for the transport, Exim sent two messages over the
4323 same TCP/IP connection, and then failed with "socket operation on non-
4324 socket" when it tried to send the third.
4326 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4328 27. The extern definition of crypt16() in expand.c was not being excluded when
4329 the OS had its own crypt16() function.
4331 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4332 as a preferred synonym for return_size_limit, both as an option and as an
4335 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4337 30. Changed the default configuration ACL to relax the local part checking rule
4338 for addresses that are not in any local domains. For these addresses,
4339 slashes and pipe symbols are allowed within local parts, but the sequence
4340 /../ is explicitly forbidden.
4342 31. SPA server authentication was not clearing the challenge buffer before
4345 32. log_message in a "warn" ACL statement was writing to the reject log as
4346 well as to the main log, which contradicts the documentation and doesn't
4347 seem right (because no rejection is happening). So I have stopped it.
4349 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4350 However, I am unable to do any testing of this.
4352 34. Fixed an infelicity in the appendfile transport. When checking directories
4353 for a mailbox, to see if any needed to be created, it was accidentally
4354 using path names with one or more superfluous leading slashes; tracing
4355 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4357 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4358 contents are added to the log line that is written for every discarded
4359 recipient. (Previously a log_message setting was ignored.)
4361 36. The ${quote: operator now quotes the string if it is empty.
4363 37. The install script runs exim in order to find its version number. If for
4364 some reason other than non-existence or emptiness, which it checks, it
4365 could not run './exim', it was installing it with an empty version number,
4366 i.e. as "exim-". This error state is now caught, and the installation is
4369 38. An argument was missing from the function that creates an error message
4370 when Exim fails to connect to the socket for saslauthd authentication.
4371 This could cause Exim to crash, or give a corrupted message.
4373 39. Added isip, isip4, and isip6 to ${if conditions.
4375 40. The ACL variables $acl_xx are now saved with the message, and can be
4376 accessed later in routers, transports, and filters.
4378 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4379 strings in the file are not string-expanded.
4381 42. If a MAIL command specified a SIZE value that was too large to fit into an
4382 int variable, the check against message_size_limit failed. Such values are
4383 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4384 day this will have to be increased, but I don't think I want to be around
4385 when emails are that large.
4392 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4393 iconv() is not standard in FreeBSD.
4395 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4396 IPv6 enabled. The observed symptom was a segmentation fault on return from
4397 the function os_common_find_running_interfaces() in src/os.c.
4399 3. In the check_special_case() function in daemon.c I had used "errno" as an
4400 argument name, which causes warnings on some systems. This was basically a
4401 typo, since it was named "eno" in the comments!
4403 4. The code that waits for the clock to tick (at a resolution of some fraction
4404 of a second) so as to ensure message-id uniqueness was always waiting for
4405 at least one whole tick, when it could have waited for less. [This is
4406 almost certainly not relevant at current processor speeds, where it is
4407 unlikely to ever wait at all. But we try to future-proof.]
4409 5. The function that sleeps for a time interval that includes fractions of a
4410 second contained a race. It did not block SIGALRM between setting the
4411 timer, and suspending (a couple of lines later). If the interval was short
4412 and the sigsuspend() was delayed until after it had expired, the suspension
4413 never ended. On busy systems this could lead to processes getting stuck for
4416 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4417 process, before it forks any delivery processes. The open lookup caching
4418 mechanism meant that the open file or database connection was passed into
4419 the delivery process. The problem was that delivery processes always tidy
4420 up cached lookup data. This could cause a problem for the next delivery
4421 process started by the queue runner, because the external queue runner
4422 process does not know about the closure. So the next delivery process
4423 still has data in the lookup cache. In the case of a file lookup, there was
4424 no problem because closing a file descriptor in a subprocess doesn't affect
4425 the parent. However, if the lookup was caching a connection to a database,
4426 the connection was closed, and the second delivery process was likely to
4427 see errors such as "PGSQL: query failed: server closed the connection
4428 unexpectedly". The problem has been fixed by closing all cached lookups
4429 in a queue runner before running a delivery process.
4431 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4432 seem to have the "const" qualifier which it has on other OS. I've
4435 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4436 *of the same type* (client or server) with the same public name that an
4437 error should be diagnosed.
4439 9. When Exim looked up a host name for an IP address, but failed to find the
4440 original IP address when looking up the host name (a safety check), it
4441 output the message "<ip address> does not match any IP for NULL", which was
4442 confusing, to say the least. The bug was that the host name should have
4443 appeared instead of "NULL".
4445 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4446 user that is built into the binary, and the -C or -D options is used, root
4447 privilege is dropped before the configuration file is read. In addition,
4448 logging is switched to stderr instead of the normal log files. If the
4449 configuration then re-defines the Exim user, the unprivileged environment
4450 is probably not what is expected, so Exim logs a panic warning message (but
4453 However, if deliver_drop_privilege is set, the unprivileged state may well
4454 be exactly what is intended, so the warning has been cut out in that case,
4455 and Exim is allowed to try to write to its normal log files.
4461 1. smtp_return_error_details was not giving details for temporary sender
4462 or receiver verification errors.
4464 2. Diagnose a configuration error if two authenticators have the same public
4467 3. Exim used not to create the message log file for a message until the first
4468 delivery attempt. This could be confusing when incoming messages were held
4469 for policy or load reasons. The message log file is now created at the time
4470 the message is received, and an initial "Received" line is written to it.
4472 4. The automatically generated man page for command line options had a minor
4473 bug that caused no ill effects; however, a more serious problem was that
4474 the procedure for building the man page automatically didn't always
4475 operate. Consequently, release 4.20 contains an out-of-date version. This
4476 shouldn't happen again.
4478 5. When building Exim with embedded Perl support, the script that builds the
4479 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4480 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4482 6. The freeze_tell option was not being used for messages that were frozen on
4483 arrival, either by an ACL or by local_scan().
4485 7. Added the smtp_incomplete_transaction log selector.
4487 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4488 was accepting AUTH without a new EHLO.
4490 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4491 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4494 10. Logging of TCP/IP connections (when configured) now happens in the main
4495 daemon process instead of the child process, so that the TCP/IP connection
4496 count is more accurate (but it can never be perfect).
4498 11. The use of "drop" in a nested ACL was not being handled correctly in the
4499 outer ACL. Now, if condition failure induced by the nested "drop" causes
4500 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4501 or "require"), the connection is dropped.
4503 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4504 that yield "discard" can now be used with an "accept" or a "discard" verb,
4505 but an error is generated for any others (because I can't see a useful way
4506 to define what should happen).
4508 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4509 are now processed in the same way as lines in the Exim configuration file.
4510 In particular, continuation lines are supported.
4512 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4514 15. Added -ti meaning -t -i.
4516 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4517 domains, and warn by logging if others are found.
4519 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4520 returned from GIFCONF when Exim is trying to find the list of interfaces on
4521 a host. The code in os.c has been modified to copy each ifreq to an aligned
4522 structure in all cases.
4524 Also, in some cases, the returned ifreq's were being copied to a 'struct
4525 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4526 means the last couple of bytes of an IPv6 address could be chopped if the
4527 ifreq contained only a normal sockaddr (14 bytes storage).
4529 18. Named domain lists were not supported in the hosts_treat_as_local option.
4530 An entry such as +xxxx was not recognized, and was treated as a literal
4533 19. Ensure that header lines added by a DATA ACL are included in the reject log
4534 if the ACL subsequently rejects the message.
4536 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4537 MD5 (which is deprecated).
4539 21. When testing a filter file using -bf, Exim was writing a message when it
4540 took the sender from a "From " line in the message, but it was not doing so
4541 when it took $return_path from a Return-Path: header line. It now does.
4543 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4544 with a valid header line field name (a series of printing characters
4545 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4547 23. Changed "disc" in the source to "disk" to conform to the documentation and
4548 the book and for uniformity.
4550 24. Ignore Sendmail's -Ooption=value command line item.
4552 25. When execve() failed while trying to run a command in a pipe transport,
4553 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4554 could be confused with a return value of 69 from the command itself. This
4555 has been changed to 127, the value the shell returns if it is asked to run
4556 a non-existent command. The wording for the related log line suggests a
4557 non-existent command as the problem.
4559 26. If received_header_text expands to an empty string, do not add a Received:
4560 header line to the message. (Well, it adds a token one on the spool, but
4561 marks it "old" so that it doesn't get used or transmitted.)
4563 27. Installed eximstats 1.28 (addition of -nt option).
4565 28. There was no check for failure on the call to getsockname() in the daemon
4566 code. This can fail if there is a shortage of resources on the system, with
4567 ENOMEM, for example. A temporary error is now given on failure.
4569 29. Contrary to the C standard, it seems that in some environments, the
4570 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4571 program. Exim now does this explicitly; it affects the formatting of
4572 timestamps using strftime().
4574 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4575 complaints. I've now initialized all the variables, to avoid this.
4577 32. Header lines added by a system filter were not being "seen" during
4578 transport-time rewrites.
4580 33. The info_callback() function passed to OpenSSL is set up with type void
4581 (*)(SSL *, int, int), as described somewhere. However, when calling the
4582 function (actually a macro) that sets it up, the type void(*)() is
4583 expected. I've put in a cast to prevent warnings from picky compilers.
4585 34. If a DNS black list lookup found a CNAME record, but there were no A
4586 records associated with the domain it pointed at, Exim crashed.
4588 35. If a DNS black list lookup returned more than one A record, Exim ignored
4589 all but the first. It now scans all returned addresses if a particular IP
4590 value is being sought. In this situation, the contents of the
4591 $dnslist_value variable are a list of all the addresses, separated by a
4594 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4595 to accept a host name and all its aliases if the forward lookup for any of
4596 them yielded the IP address of the incoming connection. Now it accepts only
4597 those names whose forward lookup yields the correct IP address. Any other
4598 names are discarded. This closes a loophole whereby a rogue DNS
4599 administrator could create reverse DNS records to break through a
4600 wildcarded host restriction in an ACL.
4602 37. If a user filter or a system filter that ran in a subprocess used any of
4603 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4604 the wrong values were passed to the pipe command ($thisaddress had the
4605 value of $0, $0 had the value of $1, etc). This bug was introduced by
4606 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4608 38. Improved the line breaking for long SMTP error messages from ACLs.
4609 Previously, if there was no break point between 40 and 75 characters, Exim
4610 left the rest of the message alone. Two changes have been made: (a) I've
4611 reduced the minimum length to 35 characters; (b) if it can't find a break
4612 point between 35 and 75 characters, it looks ahead and uses the first one
4613 that it finds. This may give the occasional overlong line, but at least the
4614 remaining text gets split now.
4616 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4617 file descriptors might be low, and that setting 1000 would always raise it.
4618 It turns out that in some environments, the limit is already over 1000 and
4619 that lowering it causes trouble. So now Exim takes care not to decrease it.
4621 40. When delivering a message, the value of $return_path is set to $sender_
4622 address at the start of routing (routers may change the value). By an
4623 oversight, this default was not being set up when an address was tested by
4624 -bt or -bv, which affected the outcome if any router or filter referred to
4627 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4628 log only when "message" or "log_message" are set. However, if one of the
4629 conditions was an address verification, or a call to a nested ACL, the
4630 messages generated by the underlying test were being passed through. This
4631 no longer happens. The underlying message is available in $acl_verify_
4632 message for both "message" and "log_message" expansions, so it can be
4633 passed through if needed.
4635 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4636 new expansion $bh_ to give the encoded byte string without charset
4637 translation. Translation happens only if iconv() is available; HAVE_ICONV
4638 indicates this at build time. HEADERS_CHARSET gives the charset to
4639 translate to; headers_charset can change it in the configuration, and
4640 "headers charset" can change it in an individual filter file.
4642 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4643 that creates RFC 2047 encoded "words" labels them as that charset instead
4644 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4645 expansion operator; (ii) when Exim creates a From: line for a local
4646 message; (iii) when a header line is rewritten to include a "phrase" part.
4648 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4649 buggy, causing it to skip the first lines of messages whose message ID
4650 ended in 'D'. This would not have bitten before Exim release 4.14, because
4651 message IDs were unlikely to end in 'D' before then. The effect was to have
4652 incorrect size information for certain domains.
4654 45. #include "config.h" was missing at the start of the crypt16.c module. This
4655 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4658 46. If there was a timeout during a "random" callout check, Exim treated it as
4659 a failure of the random address, and carried on sending RSET and the real
4660 address. If the delay was just some slowness somewhere, the response to the
4661 original RCPT would be taken as a response to RSET and so on, causing
4662 mayhem of various kinds.
4664 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4665 when I implemented it. It didn't allow for the fact that some option values
4666 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4667 the right test for positive values.
4669 48. Domain names in DNS records are case-independent. Exim always looks them up
4670 in lower case. Some resolvers return domain names in exactly the case they
4671 appear in the zone file, that is, they may contain uppercase letters. Not
4672 all resolvers do this - some return always lower case. Exim was treating a
4673 change of case by a resolver as a change of domain, similar to a widening
4674 of a domain abbreviation. This triggered its re-routing code and so it was
4675 trying to route what was effectively the same domain again. This normally
4676 caused routing to fail (because the router wouldn't handle the domain
4677 twice). Now Exim checks for this case specially, and just changes the
4678 casing of the domain that it ultimately uses when it transmits the message
4681 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4684 50. If a filter generated a file delivery with a non-absolute name (possible if
4685 no home directory exists for the router), the forbid_file option was not
4688 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4689 the existing equality matching.
4691 52. Exim was using ints instead of ino_t variables in some places where it was
4692 dealing with inode numbers.
4694 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4695 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4696 called TMPDIR, and if it finds it is different, it changes its value.
4698 54. The smtp_printf() function is now made available to local_scan() so
4699 additional output lines can be written before returning. There is also an
4700 smtp_fflush() function to enable the detection of a dropped connection.
4701 The variables smtp_input and smtp_batched_input are exported to
4704 55. Changed the default runtime configuration: the message "Unknown user"
4705 has been removed from the ACL, and instead placed on the localuser router,
4706 using the cannot_route_message feature. This means that any verification
4707 failures that generate their own messages won't get overridden. Similarly,
4708 the "Unrouteable address" message that was in the ACL for unverifiable
4709 relay addresses has also been removed.
4711 56. Added hosts_avoid_esmtp to the smtp transport.
4713 57. The exicyclog script was not checking for the esoteric option
4714 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4715 will work only if exicyclog is run under the appropriate euid.
4717 58. Following a discussion on the list, the rules by which Exim recognises line
4718 endings on incoming messages have been changed. The -dropcr and drop_cr
4719 options are now no-ops, retained only for backwards compatibility. The
4720 following line terminators are recognized: LF CRLF CR. However, special
4721 processing applies to CR:
4723 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4724 nor a local message in the state where . is a terminator.
4726 (ii) If a bare CR is encountered in a header line, an extra space is added
4727 after the line terminator so as not to end the header. The reasoning
4728 behind this is that bare CRs in header lines are most likely either
4729 to be mistakes, or people trying to play silly games.
4731 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4732 was being incorrectly given as 18 bytes larger than it should have been.
4733 This is a VOB (very old bug).
4735 60. This may never have affected anything current, but just in case it has:
4736 When the local host is found other than at the start of a list of hosts,
4737 the local host, those with the same MX, and any that follow, are discarded.
4738 When the list in question was part of a longer list of hosts, the following
4739 hosts (not currently being processed) were also being discarded. This no
4740 longer happens. I'm not sure if this situation could ever has previously
4743 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4746 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4747 header. This is something that is recommended in a new Internet Draft, and
4748 is something that is documented as being done by Sendmail. There are two
4749 possible values. For messages generated by the autoreply transport, Exim
4752 Auto-Submitted: auto-replied
4754 whereas for all other generated messages (e.g. bounces) it adds
4756 Auto-Submitted: auto-generated
4758 63. The "personal" condition in filters now includes a test for the
4759 Auto-Submitted: header. If it contains the string "auto-" the message it
4760 not considered personal.
4762 64. Added rcpt_include_affixes as a generic transport option.
4764 65. Added queue_only_override (default true).
4766 66. Added the syslog_duplication option.
4768 67. If what should have been the first header line of a message consisted of
4769 a space followed by a colon, Exim was mis-interpreting it as a header line.
4770 It isn't of course - it is syntactically invalid and should therefore be
4771 treated as the start of the message body. The misbehaviour could have
4772 caused a number of strange effects, including loss of data in subsequent
4773 header lines, and spool format errors.
4775 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4776 client host had authenticated. This control can now be exercised by an ACL
4777 for more flexibility.
4779 69. By default, callouts do not happen when testing with -bh. There is now a
4780 variant, -bhc, which does actually run the callout code, including
4781 consulting and updating the callout cache.
4783 70. Added support for saslauthd authentication, courtesy of Alexander
4786 71. If statvfs() failed on the spool or log directories while checking their
4787 size for availability, Exim confusingly gave the error "space shortage".
4788 Furthermore, in debugging mode it crashed with a floating point exception.
4789 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4790 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4791 serious problem, Exim now writes to the main and panic logs when this
4792 happens, with details of the failure. It then refuses to accept the
4793 incoming message, giving the message "spool directory problem" or "log
4794 directory problem" with a 421 code for SMTP messages.
4796 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4797 0, 1, and 2 exist, because some OS complain for execs without them (see
4798 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4799 descriptors. However, the code omitted to check that the open succeeded,
4800 causing mysterious errors if for some reason the permissions on /dev/null
4801 got screwed. Now Exim writes a message to the main and panic logs, and
4802 bombs out if it can't open /dev/null.
4804 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4805 interact so that it is all more flexible. It is supposed to remain
4806 backwards compatible. Also added extra_local_interfaces.
4808 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4809 to bomb out with an assertion failure - to the client this appears as a
4810 connection drop. This problem occurs in the part of the code that was taken
4811 from the Samba project. Fortunately, the assertion is in a very simple
4812 function, so I have fixed this by reproducing the function inline in the
4813 one place where it is called, and arranging for authentication to fail
4814 instead of killing the process with assert().
4816 75. The SPA client code was not working when the server requested OEM rather
4817 than Unicode encoding.
4819 76. Added code to make require_files with a specific uid setting more usable in
4820 the case where statting the file as root fails - usually a non-root-mounted
4821 NFS file system. When this happens and the failure is EACCES, Exim now
4822 forks a subprocess and does the per-uid checking as the relevant uid.
4824 77. Added process_log_path.
4826 78. If log_file_path was not explicitly set, a setting of check_log_space or
4827 check_log_inodes was ignored.
4829 79. If a space check for the spool or log partitions fails, the incident is now
4830 logged. Of course, in the latter case the data may get lost...
4832 80. Added the %p formatting code to string_format() so that it can be used to
4833 print addresses in debug_print(). Adjusted all the address printing in the
4834 debugging in store.c to use %p rather than %d.
4836 81. There was a concern that a line of code in smtp_in.c could overflow a
4837 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4838 initially expressed, the concern was not well-founded, because trailing
4839 spaces are removed early. However, if the trailing spaces were followed by
4840 a NULL, they did not get removed, so the overflow was possible. Two fixes
4843 (a) I re-wrote the offending code in a cleaner fashion.
4844 (b) If an incoming SMTP command contains a NULL character, it is rejected
4847 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4848 runs initgroups(), so that if the Exim user is in any additional groups,
4849 they will be used during message reception.
4855 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob