1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.529 2007/08/29 14:02:22 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
60 PH/07 The error message for a badly-placed control=no_multiline_responses left
61 "_responses" off the end of the name.
63 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
65 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
66 (without spaces) instead of just copying the configuration text.
68 PH/10 Added the /noupdate option to the ratelimit ACL condition.
70 PH/11 Added $max_received_linelength.
72 PH/12 Added +ignore_defer and +include_defer to host lists.
74 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
75 way in which PCRE > 7.0 is built.
77 PH/14 Implemented queue_only_load_latch.
79 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
80 MAIL command. The effect was to mangle the value on 64-bit systems.
82 PH/16 Another patch from the Sieve maintainer.
84 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
86 PH/18 If a system quota error occurred while trying to create the file for
87 a maildir delivery, the message "Mailbox is full" was not appended to the
88 bounce if the delivery eventually timed out. Change 4.67/27 below applied
89 only to a quota excession during the actual writing of the file.
91 PH/19 It seems that peer DN values may contain newlines (and other non-printing
92 characters?) which causes problems in log lines. The DN values are now
93 passed through string_printing() before being added to log lines.
95 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
96 and InterBase are left for another time.)
98 PH/21 Added message_body_newlines option.
100 PH/22 Guard against possible overflow in moan_check_errorcopy().
102 PH/23 POSIX allows open() to be a macro; guard against that.
108 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
109 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
112 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
113 issue a MAIL command.
115 PH/02 In an ACL statement such as
117 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
119 if a client was not listed at all, or was listed with a value other than
120 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
121 the condition was not true (as it should be), so access was not denied.
122 The bug was that the ! inversion was incorrectly passed on to the second
123 item. This has been fixed.
125 PH/03 Added additional dnslists conditions == and =& which are different from
126 = and & when the dns lookup returns more than one IP address.
128 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
129 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
131 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
132 FSYNC, which compiles an option called disable_fsync that allows for
133 bypassing fsync(). The documentation is heavily laced with warnings.
135 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
137 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
138 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
139 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
140 including adding "make clean"; (3) Added -fPIC when compiling the test
141 dynamically loaded module, to get rid of a warning.
143 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
144 message fails, move_frozen_messages = true and ignore_bounce_errors_after
145 = 0s. The bug is otherwise harmless.
147 PH/07 There was a bug in the dovecot authenticator such that the value of
148 $auth1 could be overwritten, and so not correctly preserved, after a
149 successful authentication. This usually meant that the value preserved by
150 the server_setid option was incorrect.
152 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
154 PH/09 Installed PCRE release 7.0.
156 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
157 run for batched SMTP input. It is now run at the start of every message
158 in the batch. While fixing this I discovered that the process information
159 (output by running exiwhat) was not always getting set for -bs and -bS
160 input. This is fixed, and it now also says "batched" for BSMTP.
162 PH/11 Added control=no_pipelining.
164 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
165 patch, slightly modified), and move the expansion of helo_data till after
166 the connection is made in the smtp transport (so it can use these
169 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
171 PH/14 Added log_selector = +pid.
173 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
175 PH/16 Add ${if forany and ${if forall.
177 PH/17 Added dsn_from option to vary the From: line in DSNs.
179 PH/18 Flush SMTP output before performing a callout, unless control =
180 no_callout_flush is set.
182 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
183 was true (the default) a successful delivery failed to delete the retry
184 item, thus causing premature timeout of the address. The bug is now
187 PH/20 Added hosts_avoid_pipelining to the smtp transport.
189 PH/21 Long custom messages for fakedefer and fakereject are now split up
190 into multiline reponses in the same way that messages for "deny" and
191 other ACL rejections are.
193 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
194 with slight modification.
196 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
197 draft, changing the syntax and factoring some duplicate code".
199 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
200 for deliveries of the second and subsequent messages over the same SMTP
203 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
204 ${reduce, with only minor "tidies".
206 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
208 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
209 expansion side effects.
211 PH/27 When a message times out after an over-quota error from an Exim-imposed
212 quota, the bounce message says "mailbox is full". This message was not
213 being given when it was a system quota that was exceeded. It now should
216 MH/03 Made $recipients available in local_scan(). local_scan() already has
217 better access to the recipient list through recipients_list[], but
218 $recipients can be useful in postmaster-provided expansion strings.
220 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
221 in the case of a MAIL command with additional options following the
222 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
223 were accidentally chopped off.
225 PH/29 SMTP synchronization checks are implemented when a command is read -
226 there is a check that no more input is waiting when there shouldn't be
227 any. However, for some commands, a delay in an ACL can mean that it is
228 some time before the response is written. In this time, more input might
229 arrive, invalidly. So now there are extra checks after an ACL has run for
230 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
231 pipelining has not been advertised.
233 PH/30 MH's patch to allow iscntrl() characters to be list separators.
235 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
236 returned in the SMTP response when smtp_return_error_details was false.
239 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
240 instead of the C I/O that was originally supplied, because problems were
243 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
244 Exim which did not show up earlier: it was assuming that a call to
245 SSL_CTX_set_info_callback() might give an error value. In fact, there is
246 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
247 was a macro that became an assignment, so it seemed to work. This has
248 changed to a proper function call with a void return, hence the compile
249 error. Exim's code has been fixed.
251 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
254 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
256 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
258 PH/37 If a message is not accepted after it has had an id assigned (e.g.
259 because it turns out to be too big or there is a timeout) there is no
260 "Completed" line in the log. When some messages of this type were
261 selected by exigrep, they were listed as "not completed". Others were
262 picked up by some special patterns. I have improved the selection
263 criteria to be more general.
265 PH/38 The host_find_failed option in the manualroute router can now be set
266 to "ignore", to completely ignore a host whose IP address cannot be
267 found. If all hosts are ignored, the behaviour is controlled by the new
268 host_all_ignored option.
270 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
271 homing or because of multiple MX records with /mx) generated more than
272 one IP address, and the following item turned out to be the local host,
273 all the secondary addresses of the first item were incorrectly removed
274 from the list, along with the local host and any following hosts (which
275 is what is supposed to happen).
277 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
278 whoever called Exim into the -H file. In the case of the daemon it was
279 behaving confusingly. When first started, it used values for whoever
280 started the daemon, but after a SIGHUP it used the Exim user (because it
281 calls itself on a restart). I have changed the code so that it now always
284 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
285 message are rejected with the same error (e.g. no authentication or bad
286 sender address), and a DATA command is nevertheless sent (as can happen
287 with PIPELINING or a stupid MUA), the error message that was given to the
288 RCPT commands is included in the rejection of the DATA command. This is
289 intended to be helpful for MUAs that show only the final error to their
292 PH/42 Another patch from the Sieve maintainer.
294 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
295 Eximstats - Fixed some broken HTML links and added missing column headers
297 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
298 columns for Rejects, Temp Rejects, Ham, and Spam rows.
300 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
302 PH/43 Yet another patch from the Sieve maintainer.
304 PH/44 I found a way to check for a TCP/IP connection going away before sending
305 the response to the final '.' that terminates a message, but only in the
306 case where the client has not sent further data following the '.'
307 (unfortunately, this is allowed). However, in many cases there won't be
308 any further data because there won't be any more messages to send. A call
309 to select() can be used: if it shows that the input is "ready", there is
310 either input waiting, or the socket has been closed. An attempt to read
311 the next input character can distinguish the two cases. Previously, Exim
312 would have sent an OK response which the client would never have see.
313 This could lead to message repetition. This fix should cure that, at
314 least in a lot of common cases.
316 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
317 advertised in response to EHLO.
323 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
324 fixed by 4.65/MH/01 (is this a record?) are fixed:
326 (i) An empty string was always treated as zero by the numeric comparison
327 operators. This behaviour has been restored.
329 (ii) It is documented that the numeric comparison operators always treat
330 their arguments as decimal numbers. This was broken in that numbers
331 starting with 0 were being interpreted as octal.
333 While fixing these problems I realized that there was another issue that
334 hadn't been noticed. Values of message_size_limit (both the global option
335 and the transport option) were treated as octal if they started with 0.
336 The documentation was vague. These values are now always treated as
337 decimal, and I will make that clear in the documentation.
343 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
344 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
347 MH/01 Don't check that the operands of numeric comparison operators are
348 integers when their expansion is in "skipping" mode (fixes bug
349 introduced by 4.64-PH/07).
351 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
352 child addresses, Exim now panics and dies. Previously, because the count
353 is held in a short int, deliveries were likely to be lost. As such a
354 large number of recipients for a single message is ridiculous
355 (performance will be very, very poor), I have chosen to impose a limit
356 rather than extend the field.
362 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
363 leftover -K file (the existence of which was triggered by #402).
364 While we were at it, introduced process PID as part of the -K
365 filename. This should rule out race conditions when creating
368 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
369 processing considerably. Previous code took too long for large mails,
370 triggering a timeout which in turn triggers #401.
372 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
373 in the DK code in transports.c. sendfile() is not really portable,
374 hence the _LINUX specificness.
376 TF/01 In the add_headers option to the mail command in an Exim filter,
377 there was a bug that Exim would claim a syntax error in any
378 header after the first one which had an odd number of characters
381 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
382 callout verification, Exim cached a "reject" for the entire domain. This
383 is correct for most verifications, but it is not correct for a recipient
384 verification with use_sender or use_postmaster set, because in that case
385 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
386 case of MAIL FROM:<> rejection from other early rejections (e.g.
387 rejection of HELO). When verifying a recipient using a non-null MAIL
388 address, the cache is ignored if it shows MAIL FROM:<> rejection.
389 Whatever the result of the callout, the value of the domain cache is
390 left unchanged (for any other kind of callout, getting as far as trying
391 RCPT means that the domain itself is ok).
393 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
396 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
397 manifest itself as EPIPE rather than ECONNECT. When tidying away a
398 session, the daemon ignores ECONNECT errors and logs others; it now
399 ignores EPIPE as well.
401 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
402 (quoted-printable decoding).
404 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
405 later the small subsequent patch to fix an introduced bug.
407 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
409 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
411 PH/08 An error is now given if message_size_limit is specified negative.
413 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
414 to be given (somewhat) arbitrary names.
416 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
419 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
420 miscellaneous code fixes
422 PH/10 Added the log_reject_target ACL modifier to specify where to log
425 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
426 hostname. This is wrong, because it relates to the incoming message (and
427 probably the interface on which it is arriving) and not to the outgoing
428 callout (which could be using a different interface). This has been
429 changed to use the value of the helo_data option from the smtp transport
430 instead - this is what is used when a message is actually being sent. If
431 there is no remote transport (possible with a router that sets up host
432 addresses), $smtp_active_hostname is used.
434 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
435 tweaks were necessary in order to get it to work (see also 21 below):
436 (a) The code assumed that strncpy() returns a negative number on buffer
437 overflow, which isn't the case. Replaced with Exim's string_format()
439 (b) There were several signed/unsigned issues. I just did the minimum
440 hacking in of casts. There is scope for a larger refactoring.
441 (c) The code used strcasecmp() which is not a standard C function.
442 Replaced with Exim's strcmpic() function.
443 (d) The code set only $1; it now sets $auth1 as well.
444 (e) A simple test gave the error "authentication client didn't specify
445 service in request". It would seem that Dovecot has changed its
446 interface. Fortunately there's a specification; I followed it and
447 changed what the client sends and it appears to be working now.
449 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
452 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
453 address A is aliased to B and C, where B exists and C does not. Without
454 -v the output is "A verified" because verification stops after a
455 successful redirection if more than one address is generated. However,
456 with -v the child addresses are also verified. Exim was outputting "A
457 failed to verify" and then showing the successful verification for C,
458 with its parentage. It now outputs "B failed to verify", showing B's
459 parentage before showing the successful verification of C.
461 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
462 look up a TXT record in a specific list after matching in a combined
465 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
466 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
467 they consult the DNS. I had assumed they would set it the way they
468 wanted; and indeed my experiments on Linux seem to show that in some
469 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
470 To be on the safe side, however, I have now made the interface to
471 host_find_byname() similar to host_find_bydns(), with an argument
472 containing the DNS resolver options. The host_find_byname() function now
473 sets these options at its start, just as host_find_bydns() does. The smtp
474 transport options dns_qualify_single and dns_search_parents are passed to
475 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
476 of host_find_byname() use the default settings of RES_DEFNAMES
477 (qualify_single) but not RES_DNSRCH (search_parents).
479 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
480 spool_read_header() do less string testing, by means of a preliminary
481 switch on the second character of optional "-foo" lines. (This is
482 overdue, caused by the large number of possibilities that now exist.
483 Originally there were few.) While I was there, I also converted the
484 str(n)cmp tests so they don't re-test the leading "-" and the first
485 character, in the hope this might squeeze out yet more improvement.
487 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
488 flag allowing group syntax was set by the header_syntax check but not
489 turned off, possible causing trouble later; (2) The flag was not being
490 set at all for the header_verify test, causing "group"-style headers to
491 be rejected. I have now set it in this case, and also caused header_
492 verify to ignore an empty address taken from a group. While doing this, I
493 came across some other cases where the code for allowing group syntax
494 while scanning a header line wasn't quite right (mostly, not resetting
495 the flag correctly in the right place). These bugs could have caused
496 trouble for malformed header lines. I hope it is now all correct.
498 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
499 with the "reply" argument non-NULL. The code, however (which originally
500 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
501 but it didn't always do it. This confused somebody who was copying the
502 code for some other use. I have removed all the tests.
504 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
505 feature that was used to support insecure browsers during the U.S. crypto
506 embargo. It requires special client support, and Exim is probably the
507 only MTA that supported it -- and would never use it because real RSA is
508 always available. This code has been removed, because it had the bad
509 effect of slowing Exim down by computing (never used) parameters for the
510 RSA_EXPORT functionality.
512 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
513 authenticator to fail if there's a tab character in the incoming data
514 (there should never be unless someone is messing about, as it's supposed
515 to be base64-encoded). Also added, on Timo's advice, the "secured" option
516 if the connection is using TLS or if the remote IP is the same as the
517 local IP, and the "valid-client-cert option" if a client certificate has
520 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
521 authenticators. This can be used for authorization after authentication
522 succeeds. (In the case of plaintext, it servers for both authentication
525 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
526 if any retry times were supplied.
528 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
529 connection, where there is no client IP address to check. In this
530 situation, the verify now always succeeds.
532 PH/25 Applied John Jetmore's -Mset patch.
534 PH/26 Added -bem to be like -Mset, but loading a message from a file.
536 PH/27 In a string expansion for a processed (not raw) header when multiple
537 headers of the same name were present, leading whitespace was being
538 removed from all of them, but trailing whitespace was being removed only
539 from the last one. Now trailing whitespace is removed from each header
540 before concatenation. Completely empty headers in a concatenation (as
543 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
544 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
546 PH/29 [Removed. This was a change that I later backed out, and forgot to
547 correct the ChangeLog entry (that I had efficiently created) before
548 committing the later change.]
550 PH/30 Exim was sometimes attempting to deliver messages that had suffered
551 address errors (4xx response to RCPT) over the same connection as other
552 messages routed to the same hosts. Such deliveries are always "forced",
553 so retry times are not inspected. This resulted in far too many retries
554 for the affected addresses. The effect occurred only when there were more
555 hosts than the hosts_max_try setting in the smtp transport when it had
556 the 4xx errors. Those hosts that it had tried were not added to the list
557 of hosts for which the message was waiting, so if all were tried, there
558 was no problem. Two fixes have been applied:
560 (i) If there are any address or message errors in an SMTP delivery, none
561 of the hosts (tried or untried) are now added to the list of hosts
562 for which the message is waiting, so the message should not be a
563 candidate for sending over the same connection that was used for a
564 successful delivery of some other message. This seems entirely
565 reasonable: after all the message is NOT "waiting for some host".
566 This is so "obvious" that I'm not sure why it wasn't done
567 previously. Hope I haven't missed anything, but it can't do any
568 harm, as the worst effect is to miss an optimization.
570 (ii) If, despite (i), such a delivery is accidentally attempted, the
571 routing retry time is respected, so at least it doesn't keep
572 hammering the server.
574 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
575 in ${readsocket because some servers need this prod.
577 PH/32 Added some extra debug output when updating a wait-xxx database.
579 PH/33 The hint "could be header name not terminated by colon", which has been
580 given for certain expansion errors for a long time, was not being given
581 for the ${if def:h_colon_omitted{... case.
583 PH/34 The spec says: "With one important exception, whenever a domain list is
584 being scanned, $domain contains the subject domain." There was at least
585 one case where this was not true.
587 PH/35 The error "getsockname() failed: connection reset by peer" was being
588 written to the panic log as well as the main log, but it isn't really
589 panic-worthy as it just means the connection died rather early on. I have
590 removed the panic log writing for the ECONNRESET error when getsockname()
593 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
594 runs only) independently of the message's sender address. This meant
595 that, if the 4xx error was in fact related to the sender, a different
596 message to the same recipient with a different sender could confuse
597 things. In particualar, this can happen when sending to a greylisting
598 server, but other circumstances could also provoke similar problems.
599 I have changed the default so that the retry time for these errors is now
600 based a combination of the sender and recipient addresses. This change
601 can be overridden by setting address_retry_include_sender=false in the
604 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
605 remote server are returned as part of bounce messages. This was not
606 happening for LMTP over a pipe (the lmtp transport), but now it is the
607 same for both kinds of LMTP.
609 PH/38 Despite being documented as not happening, Exim was rewriting addresses
610 in header lines that were in fact CNAMEs. This is no longer the case.
612 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
613 and queue runs started by the daemon processed all messages. This has
614 been fixed so that -R and -S can now usefully be given with -q<time>.
616 PH/40 Import PCRE release 6.7 (fixes some bugs).
618 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
620 PH/42 Give an error if -q is specified more than once.
622 PH/43 Renamed the variables $interface_address and $interface_port as
623 $received_ip_address and $received_port, to make it clear that these
624 values apply to message reception, and not to the outgoing interface when
625 a message is delivered. (The old names remain recognized, of course.)
627 PH/44 There was no timeout on the connect() call when using a Unix domain
628 socket in the ${readsocket expansion. There now is.
630 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
631 be meaningful with "accept".
633 SC/01 Eximstats V1.43
634 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
636 SC/02 Eximstats V1.44
637 Use a glob alias rather than an array ref in the generated
638 parser. This improves both readability and performance.
640 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
641 Collect SpamAssassin and rejection statistics.
642 Don't display local sender or destination tables unless
643 there is data to show.
644 Added average volumes into the top table text output.
646 SC/04 Eximstats V1.46
647 Collect data on the number of addresses (recipients)
648 as well as the number of messages.
650 SC/05 Eximstats V1.47
651 Added 'Message too big' to the list of mail rejection
652 reasons (thanks to Marco Gaiarin).
654 SC/06 Eximstats V1.48
655 Mainlog lines which have GMT offsets and are too short to
656 have a flag are now skipped.
658 SC/07 Eximstats V1.49 (Alain Williams)
659 Added the -emptyok flag.
661 SC/08 Eximstats V1.50
662 Fixes for obtaining the IP address from reject messages.
664 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
665 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
666 whitesspace changes from 4.64-PH/27
668 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
671 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
672 are found, allow negative numbers in numeric criteria)
674 JJ/06 exipick.20061117.2, added new $message_body_missing variable
676 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
677 to match changes made in 4.64-PH/43
679 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
681 PH/47 Put in an explicit test for a DNS lookup of an address record where the
682 "domain" is actually an IP address, and force a failure. This locks out
683 those revolvers/nameservers that support "A-for-A" lookups, in
684 contravention of the specifications.
686 PH/48 When a host name was looked up from an IP address, and the subsequent
687 forward lookup of the name timed out, the host name was left in
688 $sender_host_name, contrary to the specification.
690 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
691 restricted to single-key lookups, Exim was not diagnosing an error if
692 * or *@ was used with a query-style lookup.
694 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
696 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
697 long ago, but noone interested enough thought of it. Let's just say that
698 the "1.1" means that there are some new functions that weren't there at
699 some point in the past.
701 PH/51 Error processing for expansion failure of helo_data from an smtp
702 transport during callout processing was broken.
704 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
705 tested/used via the -bh/-bhc/-bs options.
707 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
708 bug, fixed in subsequent PCRE releases).
710 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
711 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
713 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
719 SC/01 Use a glob alias rather than an array ref in eximstats generated
720 parser. This improves both readability and performance.
722 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
723 Don't display local sender or destination tables in eximstats unless
724 there is data to show.
725 Added average volumes into the eximstats top table text output.
727 SC/03 Collect data on the number of addresses (recipients) as well
728 as the number of messages in eximstats.
730 TF/01 Correct an error in the documentation for the redirect router. Exim
731 does (usually) call initgroups() when daemonizing.
733 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
734 with consistent privilege compared to when running as a daemon.
736 TF/03 Note in the spec that $authenticated_id is not set for local
737 submissions from trusted users.
739 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
740 Thanks to Dean Brooks <dean@iglou.com> for the patch.
742 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
743 by adding some example configuration directives to the default
744 configuration file. A little bit of work is required to uncomment the
745 directives and define how usernames and passwords are checked, but
746 there is now a framework to start from.
748 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
749 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
750 without this. I don't know how relevant this is to other LDAP libraries.
752 PH/02 Add the verb name to the "unknown ACL verb" error.
754 PH/03 Magnus Holmgren's patch for filter_prepend_home.
756 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
758 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
759 directory not expanded when it should be if an expanded home directory
760 was set for the address (which is overridden by the transport).
762 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
765 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
766 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
767 because it is too late at that time, and has no effect.
769 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
770 security issue with \' (bugzilla #107). I could not use the
771 PQescapeStringConn() function, because it needs a PGconn value as one of
774 PH/08 When testing addresses using -bt, indicate those final addresses that
775 are duplicates that would not cause an additional delivery. At least one
776 person was confused, thinking that -bt output corresponded to deliveries.
777 (Suppressing duplicates isn't a good idea as you lose the information
778 about possibly different redirections that led to the duplicates.)
780 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
781 systems where poll() doesn't work, in particular OS X.
783 PH/10 Added more information to debugging output for retry time not reached.
785 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
786 operations in malware.c.
788 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
791 PH/13 If write_rejectlog was set false when logging was sent to syslog with
792 syslog_duplication set false, log lines that would normally be written
793 both the the main log and to the reject log were not written to syslog at
796 PH/14 In the default configuration, change the use of "message" in ACL warn
797 statements to "add_header".
799 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
800 not followed by a command (e.g. "seen endif").
802 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
803 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
806 PH/17 Added extra conditions to the default value of delay_warning_condition
810 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
811 { match{$h_precedence:}{(?i)bulk|list|junk} } \
812 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
815 The Auto-Submitted: and various List- headers are standardised, whereas I
816 don't think Precedence: ever was.
818 PH/18 Refactored debugging code in route_finduser() to show more information,
819 in particular, the error code if getpwnam() issues one.
821 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
822 This is apparently needed in addition to the PH/07 change above to avoid
823 any possible encoding problems.
825 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
826 but not after initializing Perl.
828 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
829 output them only if debugging. By default they are written stderr,
830 apparently, which is not desirable.
832 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
835 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
838 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
840 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
841 authenticated or an ident call has been made. Suppress the default
842 values for $authenticated_id and $authenticated_sender (but permit -oMai
843 and -oMas) when testing with -bh.
845 PH/24 Re-jigged the order of the tests in the default configuration so that the
846 tests for valid domains and recipients precede the DNS black list and CSA
847 tests, on the grounds that those ones are more expensive.
849 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
850 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
851 This bug exists in every version of Exim that I still have, right back to
854 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
855 However, an attempt to turn on case-sensitivity in a regex key by
856 including (?-i) didn't work because the subject string was already
857 lowercased, and the effects were non-intuitive. It turns out that a
858 one-line patch can be used to allow (?-i) to work as expected.
864 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
865 other effects) broke the use of negated acl sub-conditions.
867 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
870 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
871 "Deny" causes Exim to reject the incoming connection with a 554 error.
872 Unfortunately, if there is a major crisis, such as a disk failure,
873 tcp-wrappers gives "deny", whereas what one would like would be some
874 kind of temporary error. A kludge has been added to help with this.
875 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
876 554 error is used if errno is still zero or contains ENOENT (which occurs
877 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
880 PH/03 Add -lutil to the default FreeBSD LIBS setting.
882 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
883 errors. Otherwise a message that provokes a temporary error (when other
884 messages do not) can cause a whole host to time out.
886 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
887 addresses were routed directly to files or pipes from a redirect router.
888 File deliveries just didn't batch; pipe deliveries might have suffered
891 PH/06 A failure to get a lock for a hints database would erroneously always say
892 "Failed to get write lock", even when it was really a read lock.
894 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
895 of 0600. This has been changed to use the value of the lockfile_mode
896 option (which defaults to 0600).
898 PH/08 Applied small patch from the Sieve maintainer.
900 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
901 folder from quota calculations, a direct delivery into this folder messed
902 up the contents of the maildirsize file. This was because the regex was
903 used only to exclude .Trash (or whatever) when the size of the mailbox
904 was calculated. There was no check that a delivery was happening into an
905 excluded directory. This bug has been fixed by ignoring all quota
906 processing for deliveries into excluded directories.
908 PH/10 Added the maildirfolder_create_regex option to appendfile.
914 PH/01 The code for finding all the local interface addresses on a FreeBSD
915 system running IPv6 was broken. This may well have applied to all BSD
916 systems, as well as to others that have similar system calls. The broken
917 code found IPv4 interfaces correctly, but gave incorrect values for the
918 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
919 that it would not match correctly against @[] and not recognize the IPv6
922 PH/02 The ipliteral router was not recognizing addresses of the form user@
923 [ipv6:....] because it didn't know about the "ipv6:" prefix.
925 PH/03 Added disable_ipv6.
927 PH/04 Changed $reply_address to use the raw form of the headers instead of the
928 decoded form, because it is most often used to construct To: headers
929 lines in autoreplies, and the decoded form may well be syntactically
930 invalid. However, $reply_address has leading white space removed, and all
931 newlines turned into spaces so that the autoreply transport does not
934 PH/05 If group was specified without a user on a router, and no group or user
935 was specified on a transport, the group from the router was ignored.
937 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
938 for visible compile-time settings that can be used to change these
939 numbers, for those that want even more. Backwards compatibility with old
940 spool files has been maintained. However, going back to a previous Exim
941 release will lost any variables that are in spool files.
943 PH/07 Two small changes when running in the test harness: increase delay when
944 passing a TCP/IP connection to a new process, in case the original
945 process has to generate a bounce, and remove special handling of
946 127.0.0.2 (sic), which is no longer necessary.
948 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
949 be the same on different OS.
951 PH/09 Moved a debug statement in filter processing to avoid a race problem when
954 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
955 whether --show-vars was specified or not
957 JJ/02 exipick: Added support for new ACL variable spool format introduced
960 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
961 syntactically invalid From: or Reply-to: line, and a filter used this to
962 generate an autoreply, and therefore failed to obtain an address for the
963 autoreply, Exim could try to deliver to a non-existent relative file
964 name, causing unrelated and misleading errors. What now happens is that
965 it logs this as a hard delivery error, but does not attempt to create a
968 PH/11 The exinext utility has a -C option for testing purposes, but although
969 the given file was scanned by exinext itself; it wasn't being passed on
970 when Exim was called.
972 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
973 an end-of-file indication when reading a command response.
975 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
976 compiled. In many other places in Exim, IPv6 addresses are always
977 recognized, so I have changed this. It also means that IPv4 domain
978 literals of the form [IPV4:n.n.n.n] are now always recognized.
980 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
981 used if the router is not running as root, for example, when verifying at
982 ACL time, or when using -bh. The debugging output from this situation was
983 non-existent - all you got was a failure to exec. I have made two
986 (a) Failures to set uid/gid, the current directory, or a process leader
987 in a subprocess such as that created by queryprogram now generate
988 suitable debugging ouput when -d is set.
990 (b) The queryprogram router detects when it is not running as root,
991 outputs suitable debugging information if -d is set, and then runs
992 the subprocess without attempting to change uid/gid.
994 PH/15 Minor change to Makefile for building test_host (undocumented testing
997 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
998 additional section of a DNS packet that returns MX or SRV records.
999 Instead, it always explicitly searches for A/AAAA records. This avoids
1000 major problems that occur when a DNS server includes only records of one
1001 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1002 fixed another bug: if SRV records were looked up and the corresponding
1003 address records were *not* found in the additional section, the port
1004 values from the SRV records were lost.
1006 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1007 using the correct key (the original address) when searching the retry
1008 rules in order to find which one to use for generating the retry hint.
1010 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1011 adding the default one. Similarly, if it contains a Reply-To: header, the
1012 errors_reply_to option, if set, is not used.
1014 PH/19 When calculating a retry time, Exim used to measure the "time since
1015 failure" by looking at the "first failed" field in the retry record. Now
1016 it does not use this if it is later than than the arrival time of the
1017 message. Instead it uses the arrival time. This makes for better
1018 behaviour in cases where some deliveries succeed, thus re-setting the
1019 "first failed" field. An example is a quota failure for a huge message
1020 when small messages continue to be delivered. Without this change, the
1021 "time since failure" will always be short, possible causing more frequent
1022 delivery attempts for the huge message than are intended.
1023 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1025 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1026 $1, $2, $3) because the numerical variables can be reset during some
1027 expansion items (e.g. "match"), thereby losing the authentication data.
1029 PH/21 Make -bV show the size of off_t variables so that the test suite can
1030 decide whether to run tests for quotas > 2G.
1032 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1033 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1034 filecount value is greater than 2G or if a quota value is greater than 2G
1035 on a system where the size of off_t is not greater than 4, a panic error
1038 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1039 never match. The debug and -bh output now contains an explicit error
1040 message indicating a malformed IPv4 address or mask.
1042 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1043 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1044 PH/23 above applies.
1046 PH/25 Do not write to syslog when running in the test harness. The only
1047 occasion when this arises is a failure to open the main or panic logs
1048 (for which there is an explicit test).
1050 PH/26 Added the /no_tell option to "control=freeze".
1052 PH/27 If a host name lookup failed very early in a connection, for example, if
1053 the IP address matched host_lookup and the reverse lookup yielded a name
1054 that did not have a forward lookup, an error message of the form "no IP
1055 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1056 could be logged. Now it outputs the IP address instead of "NULL".
1058 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1059 allows the sender and the authenticated sender to be set when
1060 submitting a message from within Exim. Since child_open_exim() is
1061 documented for local_scan(), the new function should be too.
1063 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1064 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1065 results in an empty string is now treated as unset.
1067 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1069 PH/31 Added sender_verify_fail logging option.
1071 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1072 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1073 tidied the source and removed it altogether.
1075 PH/33 When a queue run was abandoned because the load average was too high, a
1076 log line was always written; now it is written only if the queue_run log
1077 selector is set. In addition, the log line for abandonment now contains
1078 information about the queue run such as the pid. This is always present
1079 in "start" and "stop" lines but was omitted from the "abandon" line.
1081 PH/34 Omit spaces between a header name and the colon in the error message that
1082 is given when verify = headers_syntax fails (if there are lots of them,
1083 the message gets confusing).
1085 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1086 names, as there are now some PTR records that contain slashes. This check
1087 is only to protect against broken name servers that fall over on strange
1088 characters, so the fact that it applies to all lookups doesn't matter.
1090 PH/36 Now that the new test suite is complete, we can remove some of the
1091 special code in Exim that was needed for the old test suite. For example,
1092 sorting DNS records because real resolvers return them in an arbitrary
1093 order. The new test suite's fake resolver always returns records in the
1096 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1097 bounces) except when queue_only is set, to avoid logging races between
1098 the different processes.
1100 PH/38 Panic-die if .include specifies a non-absolute path.
1102 PH/39 A tweak to the "H" retry rule from its user.
1104 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1105 a label. They prevented compilation on older perls.
1107 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1108 a warning to be raised on newish perls.
1110 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1111 on queue. Changes to match documented behaviour of showing count of
1112 messages matching specified criteria.
1114 PH/40 Changed the default ident timeout from 30s to 5s.
1116 PH/41 Added support for the use of login_cap features, on those BSD systems
1117 that have them, for controlling the resources used by pipe deliveries.
1119 PH/42 The content-scanning code uses fopen() to create files in which to put
1120 message data. Previously it was not paying any attention to the mode of
1121 the files. Exim runs with umask(0) because the rest of the code creates
1122 files with open(), and sets the required mode explicitly. Thus, these
1123 files were ending up world-writeable. This was not a big issue, because,
1124 being within the spool directory, they were not world-accessible. I have
1125 created a function called modefopen, which takes an additional mode
1126 argument. It sets umask(777), creates the file, chmods it to the required
1127 mode, then resets the umask. All the relevant calls to fopen() in the
1128 content scanning code have been changed to use this function.
1130 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1131 to 24 hours. This avoids potential overflow problems when processing G
1132 and H retry rules. I suspect nobody ever tinkers with this value.
1134 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1136 PH/45 When the plaintext authenticator is running as a client, the server's
1137 challenges are checked to ensure they are valid base64 strings. By
1138 default, the authentication attempt is cancelled if an invalid string is
1139 received. Setting client_ignore_invalid_base64 true ignores these errors.
1140 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1141 they are received. Thus, the responses can be made to depend on the
1142 challenges. If an invalid string is ignored, an empty string is placed in
1145 PH/46 Messages that are created by the autoreply transport now contains a
1146 References: header, in accordance with RFCs 2822 and 3834.
1148 PH/47 Added authenticated_sender_force to the smtp transport.
1150 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1152 PH/49 Installed latest patch from the Sieve maintainer.
1154 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1155 was also set, the appendfile transport was unnecessarily scanning a
1156 directory of message files (e.g. for maildir delivery) to find the count
1157 of files (along with the size), even though it did not need this
1158 information. It now does the scan only if it needs to find either the
1159 size of the count of files.
1161 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1163 PH/52 Two bugs concerned with error handling when the smtp transport is
1166 (i) Exim was not creating retry information for temporary errors given
1167 for individual recipients after the DATA command when the smtp transport
1168 was used in LMTP mode. This meant that they could be retried too
1169 frequently, and not timed out correctly.
1171 (ii) Exim was setting the flag that allows error details to be returned
1172 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1173 recipients that were returned after the DATA command.
1175 PH/53 This is related to PH/52, but is more general: for any failing address,
1176 when detailed error information was permitted to be returned to the
1177 sender, but the error was temporary, then after the final timeout, only
1178 "retry timeout exceeded" was returned. Now it returns the full error as
1179 well as "retry timeout exceeded".
1181 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1182 do this, and (what is worse) MTAs that accept it.
1184 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1185 will now be deprecated.
1187 PH/56 New os.c-cygwin from the Cygwin maintainer.
1189 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1190 formats (previously only available in exim formats via -bpr, -bpru,
1191 and -bpra. Now also available in native and exiqgrep formats)
1193 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1194 with very large, slow to parse queues
1196 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1198 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1200 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1201 responses to authentication challenges, though it was showing the
1202 challenges; (ii) I've removed the CR characters from the debug output for
1205 PH/58 Allow for the insertion of a newline as well as a space when a string
1206 is turned into more than one encoded-word during RFC 2047 encoding. The
1207 Sieve code now uses this.
1209 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1210 data_4xx, lost_connection, tls_required.
1212 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1213 message was being sent as an SMTP response.
1215 PH/61 Add -l and -k options to exicyclog.
1217 PH/62 When verifying, if an address was redirected to one new address, so that
1218 verification continued, and the new address failed or deferred after
1219 having set something in $address_data, the value of $address_data was not
1220 passed back to the ACL. This was different to the case when no
1221 redirection occurred. The value is now passed back in both cases.
1223 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1224 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1225 use login_cap.h, so on its own it isn't the distinguishing feature. The
1226 new name refers directly to the setclassresources() function.
1228 PH/65 Added configuration files for NetBSD3.
1230 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1232 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1233 is preferred over IPv4.
1235 PH/68 The bounce_return_message and bounce_return_body options were not being
1236 honoured for bounces generated during the reception of non-SMTP messages.
1237 In particular, this applied to messages rejected by the ACL. This bug has
1238 been fixed. However, if bounce_return_message is true and bounce_return_
1239 body is false, the headers that are returned for a non-SMTP message
1240 include only those that have been read before the error was detected.
1241 (In the case of an ACL rejection, they have all been read.)
1243 PH/69 The HTML version of the specification is now built in a directory called
1244 spec_html instead of spec.html, because the latter looks like a path with
1245 a MIME-type, and this confuses some software.
1247 PH/70 Catch two compiler warnings in sieve.c.
1249 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1250 function verify_get_ident() calls ip_connect() to connect a socket, but
1251 if the "connect()" function timed out, ip_connect() used to close the
1252 socket. However, verify_get_ident() also closes the socket later, and in
1253 between Exim writes to the log, which may get opened at this point. When
1254 the socket was closed in ip_connect(), the log could get the same file
1255 descriptor number as the socket. This naturally causes chaos. The fix is
1256 not to close the socket in ip_connect(); the socket should be closed by
1257 the function that creates it. There was only one place in the code where
1258 this was missing, in the iplookup router, which I don't think anybody now
1259 uses, but I've fixed it anyway.
1261 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1262 well as to direct DNS lookups. Otherwise the handling of names in host
1263 lists is inconsistent and therefore confusing.
1269 PH/01 Two changes to the default runtime configuration:
1271 (1) Move the checks for relay_from_hosts and authenticated clients from
1272 after to before the (commented out) DNS black list checks.
1274 (2) Add control=submission to the relay_from_hosts and authenticated
1275 clients checks, on the grounds that messages accepted by these
1276 statements are most likely to be submissions.
1278 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1280 (1) Generate an error if the third argument for the ${prvs expansion is
1283 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1286 (3) Reset the variables that are obtained from the first argument of
1287 ${prvscheck and used in the second argument before leaving the code,
1288 because their memory is reclaimed, so using them afterwards may do
1291 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1292 one (it's much easier than Tom thought :-).
1294 (5) Because of (4), we can now allow for the use of $prvscheck_result
1295 inside the third argument.
1297 PH/03 For some reason, the default setting of PATH when running a command from
1298 a pipe transport was just "/usr/bin". I have changed it to
1301 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1302 anything to be listed in the output from -bV.
1304 PH/05 When a filter generated an autoreply, the entire To: header line was
1305 quoted in the delivery log line, like this:
1307 => >A.N.Other <ano@some.domain> <original@ddress> ...
1309 This has been changed so that it extracts the operative address. There
1310 may be more than one such address. If so, they are comma-separated, like
1313 => >ano@some.domain,ona@other.domain <original@ddress> ...
1315 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1316 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1317 not being looked up in the rDNS to get a host name, Exim was showing the
1318 IP address twice in Received: lines, even though the IP addresses were
1319 identical. For example:
1321 Received: from [1.2.3.4] (helo=[1.2.3.4])
1323 However, if the real host name was known, it was omitting the HELO data
1324 if it matched the actual IP address. This has been tidied up so that it
1325 doesn't show the same IP address twice.
1327 PH/07 When both +timestamp and +memory debugging was on, the value given by
1328 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1329 called by the debug printing, thereby overwriting the timestamp buffer.
1330 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1332 PH/08 When the original message was included in an autoreply transport, it
1333 always said "this is a copy of the message, including all the headers",
1334 even if body_only or headers_only was set. It now gives an appropriate
1337 PH/09 Applied a patch from the Sieve maintainer which:
1339 o fixes some comments
1340 o adds the (disabled) notify extension core
1341 o adds some debug output for the result of if/elsif tests
1342 o points to the current vacation draft in the documentation
1343 and documents the missing references header update
1347 o fixes a bug in processing the envelope test (when testing
1348 multiple envelope elements, the last element determinted the
1351 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1352 Electronic Mail") by including:
1354 Auto-submitted: auto-generated
1356 in the messages that it generates (bounce messages and others, such as
1357 warnings). In the case of bounce messages for non-SMTP mesages, there was
1358 also a typo: it was using "Auto_submitted" (underscore instead of
1359 hyphen). Since every message generated by Exim is necessarily in response
1360 to another message, thes have all been changed to:
1362 Auto-Submitted: auto-replied
1364 in accordance with these statements in the RFC:
1366 The auto-replied keyword:
1368 - SHOULD be used on messages sent in direct response to another
1369 message by an automatic process,
1371 - MUST NOT be used on manually-generated messages,
1373 - MAY be used on Delivery Status Notifications (DSNs) and Message
1374 Disposition Notifications (MDNs),
1376 - MUST NOT be used on messages generated by automatic or periodic
1377 processes, except for messages which are automatic responses to
1380 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1381 to the default Received: header definition.
1383 PH/12 Added log selector acl_warn_skipped (default on).
1385 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1386 variables because (a) they are in the wrong storage pool and (b) even if
1387 they were copied, it wouldn't work properly because of the caching.
1389 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1390 checking when decoding. Apparently there are clients that generate
1391 overlong encoded strings. Why am I not surprised?
1393 PH/15 If the first argument of "${if match_address" was not empty, but did not
1394 contain an "@" character, Exim crashed. Now it writes a panic log message
1395 and treats the condition as false.
1397 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1399 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1400 extension "envelope-auth". The code is finished and in agreement with
1401 other implementations, but there is no documentation so far and in fact,
1402 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1403 not changing the active code.
1405 Print executed "if" and "elsif" statements when debugging is used. This
1406 helps a great deal to understand what a filter does.
1408 Document more things not specified clearly in RFC3028. I had all this
1409 sorted out, when out of a sudden new issues came to my mind. Oops."
1411 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1414 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1415 canonical form (as documented). However, after a host name lookup from
1416 the IP address, check_host() was doing a simple string comparison with
1417 addresses acquired from the DNS when checking that the found name did
1418 have the original IP as one of its addresses. Since any found IPv6
1419 addresses are likely to be in abbreviated form, the comparison could
1420 fail. Luckily, there already exists a function for doing the comparison
1421 by converting both addresses to binary, so now that is used instead of
1422 the text comparison.
1424 PH/20 There was another similar case to PH/19, when a complete host name was
1425 given in a host list; looking up its IP address could give an abbreviated
1426 form, whereas the current host's name might or might not be abbreviated.
1427 The same fix has been applied.
1433 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1434 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1437 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1438 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1440 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1442 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1443 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1444 Also, for "Auto-Submitted", treat anything other than "no" as
1445 non-personal, in accordance with RFC 3834. (Previously it treated
1446 anything starting "auto-" as non-personal.)
1448 TF/01 The control=submission/name=... option had a problem with syntax
1449 errors if the name included a slash character. The /name= option
1450 now slurps the rest of the string, so it can include any characters
1451 but it must come last in the list of options (after /sender_retain
1454 PH/05 Some modifications to the interface to the fake nameserver for the new
1462 TK/01 Added the "success_on_redirect" address verification option. See
1463 NewStuff for rationale and an example.
1465 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1467 PH/02 Patch to exigrep to allow it to work on syslog lines.
1469 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1470 fread() to skip over the body file's header line, because in Cygwin the
1471 header line is locked and is inaccessible.
1473 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1474 co-exist for some time) to make it clear that it is the Exim ID that is
1475 referenced, not the Message-ID: header line.
1477 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1478 string_format() function, because snprintf() does not exist on all
1481 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1482 ${stat: expansion item.
1484 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1485 protocol synchronization error", to keep the pedants happy.
1487 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1488 well as for IRIX systems, when gcc is being used. See the host.c source
1491 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1493 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1496 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1497 transport and to the smtp transport in LMTP mode.
1499 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1501 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1502 run a filter in a subprocess. This could lead to confusion in subsequent
1503 lookups in the parent process. There should also be a search_tidyup() at
1504 the end of the subprocess.
1506 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1507 only if the host matched helo_try_verify_hosts, which caused the
1508 verification to occur when the EHLO/HELO command was issued. The ACL just
1509 tested the remembered result. Now, if a previous verification attempt has
1510 not happened, "verify = helo" does it there and then.
1512 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1514 TK/03 Fix log output including CR from clamd.
1516 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1517 exist provoked a memory error which could cause a segfault.
1519 PH/15 Installed PCRE 6.2
1521 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1523 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1524 of the problem. Specifically, suggested +O2 rather than +O1 for the
1527 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1529 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1530 record, the port was not correctly logged when the outgoing_port log
1531 selector was set (it logged the transort's default port).
1533 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1534 fallback_hosts, and "hosts" in the smtp transport.
1536 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1537 host errors such as "Connection refused".
1539 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1540 authentication with radiusclient 0.4.9:
1542 - Error returned from rc_read_config was caught wrongly
1543 - Username/password not passed on to radius server due to wrong length.
1545 The presumption is that some radiusclient API changes for 4.51/PH/17
1546 were not taken care of correctly. The code is still untested by me (my
1547 Linux distribution still has 0.3.2 of radiusclient), but it was
1548 contributed by a Radius user.
1550 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1551 expanding the "port" option of the smtp transport.
1553 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1554 while reading a MIME header. Thanks to Tom Hughes for a patch.
1556 PH/24 Include config.h inside local_scan.h so that configuration settings are
1559 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1560 that in an ACL for RCPT (for example), you can examine exactly what was
1563 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1564 commands, but it was not correctly comparing the address with the actual
1565 client host address. Thus, it would show the EHLO address in Received:
1566 header lines when this was not necessary.
1568 PH/27 Added the % operator to ${eval:}.
1570 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1571 it should be ignoring failures (because with -C, for example, it has lost
1572 privilege). It wasn't ignoring creation failures other than "already
1575 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1578 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1579 because an input line was too long, either on its own, or by virtue of
1580 too many continuations, the temporary file was not being removed, and the
1581 return code was incorrect.
1583 PH/31 Missing "BOOL" in function definition in filtertest.c.
1585 PH/32 Applied Sieve patches from the maintainer.
1587 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1589 PH/33 Added "verify = not_blind".
1591 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1592 Local/Makefile (with some defaults set). These are used in built scripts
1593 such as exicyclog, but they have never been used in the exim_install
1594 script (though there are many overriding facilities there). I have
1595 arranged that the exim_install script now takes note of these two
1598 PH/35 Installed configuration files for Dragonfly.
1600 PH/36 When a locally submitted message by a trusted user did not contain a
1601 From: header, and the sender address was obtained from -f or from an SMTP
1602 MAIL command, and the trusted user did not use -F to supply a sender
1603 name, $originator_name was incorrectly used when constructing a From:
1604 header. Furthermore, $originator_name was used for submission mode
1605 messages from external hosts without From: headers in a similar way,
1606 which is clearly wrong.
1608 PH/37 Added control=suppress_local_fixups.
1610 PH/38 When log_selector = +received_sender was set, and the addition of the
1611 sender made the log line's construction buffer exactly full, or one byte
1612 less than full, an overflow happened when the terminating "\n" was
1615 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1616 when the result of a list match is failure because a DNS lookup failed.
1618 PH/40 RM_COMMAND is now used in the building process.
1620 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1621 the "build-* directories that it finds.
1623 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1624 address in a domain literal was a prefix of an interface address.
1626 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1627 when verifying a sender address, unless rewrite_headers is false.
1629 PH/44 Wrote a long comment about why errors_to addresses are verified as
1630 recipients, not senders.
1632 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1633 the ratelimit ACL was added.
1635 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1637 PH/46 Added extra information about PostgreSQL errors to the error string.
1639 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1640 avoiding the need to install special zones in a real server. This is
1641 backwards compatible; if it can't find the fake resolver, it drops back.
1642 Thus, both old and new test suites can be run.
1644 TF/02 Added util/ratelimit.pl
1646 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1647 clock is set back in time.
1649 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1650 Candler <B.Candler@pobox.com>.
1652 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1653 OK for addresses that are the result of redirections.
1655 PH/48 A number of further additions for the benefit of the new test suite,
1656 including a fake gethostbyname() that interfaces to the fake DNS resolver
1659 TF/06 The fix for widen_domains has also been applied to qualify_single and
1660 search_parents which are the other dnslookup options that can cause
1663 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1666 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1668 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1669 service" if it encountered an MX record with an empty target hostname.
1670 The message is now "an MX or SRV record indicated no SMTP service".
1672 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1673 if the DNS of the sending site is misconfigured. This is quite a
1674 common situation. This change restores the behaviour of treating a
1675 helo verification defer as a failure.
1677 PH/51 If self=fail was set on a router, the bounce message did not include the
1678 actual error message.
1684 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1686 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1687 command itself ended in error, the underlying message about the transport
1688 filter timeout was being overwritten with the pipe command error. Now the
1689 underlying error message should be appended to the second error message.
1691 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1693 PH/02 Reduce the amount of output that "make" produces by default. Full output
1694 can still be requested.
1696 PH/03 The warning log line about a condition test deferring for a "warn" verb
1697 was being output only once per connection, rather than after each
1698 occurrence (because it was using the same function as for successful
1699 "warn" verbs). This seems wrong, so I have changed it.
1701 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1702 it should not have, which might have caused a crash in the right
1703 circumstances, but probably never did.
1705 PH/04 Installed a modified version of Tony Finch's patch to make submission
1706 mode fix the return path as well as the Sender: header line, and to
1707 add a /name= option so that you can make the user's friendly name appear
1710 TF/03 Added the control = fakedefer ACL modifier.
1712 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1713 Mark Lowes for thorough testing.
1715 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1717 TK/03 Merged latest SRS patch from Miles Wilton.
1719 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1720 with the definition in sysexits.h (which is #included earlier).
1721 Fortunately, Exim does not actually use EX_OK. The code used to try to
1722 preserve the sysexits.h value, by assumimg that macro definitions were
1723 scanned for macro replacements. I have been disabused of this notion,
1724 so now the code just undefines EX_OK before #including unistd.h.
1726 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1727 in the smtp transport. When a block could not be written in a single
1728 write() function, the timeout was being re-applied to each part-write.
1729 This seems wrong - if the receiver was accepting one byte at a time it
1730 would take for ever. The timeout is now adjusted when this happens. It
1731 doesn't have to be particularly precise.
1733 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1734 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1736 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1737 without a domain if the check to <postmaster@domain> fails.
1739 SC/01 Eximstats: added -xls and the ability to specify output files
1740 (patch written by Frank Heydlauf).
1742 SC/02 Eximstats: use FileHandles for outputing results.
1744 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1746 SC/04 Eximstats: fixed display of large numbers with -nvr option
1748 SC/05 Eximstats: fixed merging of reports with empty tables.
1750 SC/06 Eximstats: added the -include_original_destination flag
1752 SC/07 Eximstats: removed tabs and trailing whitespace.
1754 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1756 TK/06 MBOX spool code: Add real "From " MBOX separator line
1757 so the .eml file is really in mbox format (even though
1758 most programs do not really care). Patch from Alex Miller.
1760 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1761 The latter is generated from $received_to and is only set if the
1762 message has one envelope recipient. SA can use these headers,
1763 obviously out-of-the-box. Patch from Alex Miller.
1765 PH/08 The ${def test on a variable was returning false if the variable's
1766 value was "0", contrary to what the specification has always said!
1767 The result should be true unless the variable is empty.
1769 PH/09 The syntax error of a character other than { following "${if
1770 def:variable_name" (after optional whitespace) was not being diagnosed.
1771 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1772 accidental colon was present, for example, could give incorrect results.
1774 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1775 result is used (not including appendfile, where other changes are about
1778 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1779 This involved changing a lot of size variables from int to off_t. It
1780 should work with maildirs and everything.
1782 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1783 spamd dying while we are connected to it.
1785 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1788 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1789 function for BDB 4.3.
1791 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1793 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1794 the sources for PCRE have been re-arranged and more files are now
1797 PH/15 The code I had for printing potentially long long variables in PH/11
1798 above was not the best (it lost precision). The length of off_t variables
1799 is now inspected at build time, and an appropriate printing format (%ld
1800 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1801 to be "long long int" or "long int". This is needed for the internal
1802 formatting function string_vformat().
1804 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1805 the configuration file to be ":syslog", then the script "guesses" where
1806 the logs files are, rather than using the compiled in default. In our
1807 case the guess is not the same as the compiled default, so the script
1808 suddenly stopped working when I started to use syslog. The patch checks
1809 to see if log_file_path is "". If so, it attempts to read it from exim
1810 with no configuration file to get the compiled in version, before it
1811 falls back to the previous guessing code."
1813 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1814 implementing BATV in an Exim configuration. See NewStuff for the gory
1817 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1818 Makefile that are specific to HP-UX.
1820 PH/18 If the "use_postmaster" option was set for a recipient callout together
1821 with the "random" option, the postmaster address was used as the MAIL
1822 FROM address for the random test, but not for the subsequent recipient
1823 test. It is now used for both.
1825 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1826 patch removes a few documentation additions to RFC 3028, because the
1827 latest draft now contains them. It adds the new en;ascii-case comparator
1828 and a new error check for 8bit text in MIME parts. Comparator and
1829 require names are now matched exactly. I enabled the subaddress
1830 extension, but it is not well tested yet (read: it works for me)."
1832 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1833 rework some of the code of TK/09 above to avoid the hardwired use of
1834 "%lld" and "long long". Replaced the call to snprintf() with a call to
1837 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1838 records point to non-existent hosts", "retry timeout exceeded", and
1839 "retry time not reached for any host after a long failure period".
1841 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1842 experimental DomainKeys support:
1844 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1845 (2) On an error such as an illegally used "control", the wrong name for
1846 the control was given.
1848 These problems did NOT occur unless DomainKeys support was compiled.
1850 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1852 PH/24 Added ${if match_ip condition.
1854 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1855 databases so that it will be absolutely obvious if a crash occurs in the
1856 DB library. This is a regular occurrence (often caused by mis-matched
1859 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1860 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1861 picked up on a user's system that detects such things. There doesn't seem
1862 to be a gcc warning option for this - only an attribute that has to be
1863 put on the function's prototype. It seems that in Fedora Core 4 they have
1864 set this on a number of new functions. No doubt there will be more in due
1867 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1868 specify a transport. However, if an address that was verified by such a
1869 router was the subject of a callout, Exim crashed because it tried to
1870 read the rcpt_include_affixes from the non-existent transport. Now it
1871 just assumes that the setting of that option is false. This bug was
1872 introduced by 4.51/PH/31.
1874 PH/28 Changed -d+all to exclude +memory, because that information is very
1875 rarely of interest, but it makes the output a lot bigger. People tend to
1876 do -d+all out of habit.
1878 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1879 code in os-type was giving problems when libc.so lives in lib64, like on
1882 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1883 aren't the modern standard, and it seems that some systems' include files
1884 don't always have them. Exim was already checking for some of the newer
1885 ones like T_AAAA, and defining it itself. I've added checks for all the
1886 record types that Exim uses.
1888 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1889 not automatically generating a new one, as it is supposed to. This
1890 prevented TLS from working. If the file did exist, but contained invalid
1891 data, a new version was generated, as expected. It was only the case of a
1892 non-existent file that was broken.
1894 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1895 with a change in libdomainkeys > 0.64.
1897 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1898 from DNS. If the selector record carries the flag, it now has
1899 precedence over the domain-wide flag.
1901 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1903 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1904 the use of a port name that isn't defined in /etc/services) occurred, the
1905 message was deferred as in a normal delivery, and thus remained on the
1906 spool, instead of being failed because of the mua_wrapper setting. This
1907 is now fixed, and I tidied up some of the mua_wrapper messages at the
1910 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1911 the messages in a hash of arrays rather than using individual hashes.
1912 This is a bit cleaner and results in dramatic memory savings, albeit
1913 at a slight CPU cost.
1915 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1916 as requested by Marc Sherman.
1918 SC/10 Eximstats: added histograms for user specified patterns as requested
1921 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1923 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1924 fopen() in the content-scanning modules that did not already have it.
1930 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1931 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1933 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1935 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1937 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1938 to test Sieve filters that use "vacation".
1940 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1941 that changes the way the GnuTLS parameters are stored in the cache file.
1942 The new format can be generated externally. For backward compatibility,
1943 if the data in the cache doesn't make sense, Exim assumes it has read an
1944 old-format file, and it generates new data and writes a new file. This
1945 means that you can't go back to an older release without removing the
1948 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1949 work if there are any delivery delays because "one_time" forces the
1950 parent to be marked "delivered", so its unseen clone is never tried
1951 again. For this reason, Exim now forbids the simultaneous setting of
1954 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1955 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1956 entry if you want to know the details. The fix, however, neglected to
1957 consider the case when local delivery batching is involved. The test for
1958 "previously delivered" was not happening when checking to see if an
1959 address could be batched with a previous (undelivered) one; under
1960 certain circumstances this could lead to multiple deliveries to the same
1963 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1964 in its include files, and this causes problems building Exim.
1966 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1967 header_syntax) but Exim was just ignoring anything given after a slash.
1968 In particular, this caused confusion with an attempt to use "verify =
1969 reverse_host_lookup/defer_ok". An error is now given when options are
1970 supplied for verify items that do not have them. (Maybe reverse_host_
1971 lookup should have a defer_ok option, but that's a different point.)
1973 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1974 defined by RFC 821) to 2048, because there were problems with some AUTH
1975 commands, and RFC 1869 says the size should be increased for extended
1976 SMTP commands that take arguments.
1978 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1981 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1982 "unknown" error; now it says that the functionality isn't in the binary.
1984 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1985 an address' error message when a string expansion fails (syntax or
1986 whatever). Otherwise the password may appear in the log. Following change
1987 PH/42 below, there is no longer a chance of it appearing in a bounce
1990 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1992 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1993 of its addresses was ever tried. (Bugzilla bug #2.)
1995 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1996 the result incorrectly in the debug output. (It correctly added a newline
1997 to what was transported.)
1999 TF/01 Added $received_time.
2001 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2002 commented out examples of how to interface to a virus scanner and to
2003 SpamAssassin. Also added commented examples of av_scanner and
2004 spamd_address settings.
2006 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2007 and controls are allowed in which ACLs. There were a couple of minor
2008 errors. Some of the entries in the conditions table (which is a table of
2009 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2010 negation of where the condition IS allowed.
2012 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2014 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2015 header file does not have a version number, so I've had to invent a new
2016 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2017 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2018 radiusclient), but it was contributed by a Radius user.
2020 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2021 files or directories, for OpenSSL.
2023 PH/19 When an Exim process that is running as root has to create an Exim log
2024 file, it does so in a subprocess that runs as exim:exim so as to get the
2025 ownership right at creation (otherwise, other Exim processes might see
2026 the file with the wrong ownership). There was no test for failure of this
2027 fork() call, which would lead to the process getting stuck as it waited
2028 for a non-existent subprocess. Forks do occasionally fail when resources
2029 run out. I reviewed all the other calls to fork(); they all seem to check
2032 PH/20 When checking for unexpected SMTP input at connect time (before writing
2033 the banner), Exim was not dealing correctly with a non-positive return
2034 from the read() function. If the client had disconnected by this time,
2035 the result was a log entry for a synchronization error with an empty
2036 string after "input=" when read() returned zero. If read() returned -1
2037 (an event I could not check), uninitialized data bytes were printed.
2038 There were reports of junk text (parts of files, etc) appearing after
2041 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2043 PH/22 Added support for macro redefinition, and (re)definition in between
2044 driver and ACL definitions.
2046 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2047 forgetting to use the resulting value; it was using the unexpanded value.
2049 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2050 hadn't been configured. The fix is from Juergen Kreileder, who
2051 understands it better than I do:
2053 "Here's what I see happening with three configured cyrus_sasl
2054 authenticators configured (plain, login, cram-md5):
2056 On startup auth_cyrus_sasl_init() gets called for each of these.
2057 This means three calls to sasl_listmech() without a specified mech_list.
2058 => SASL tests which mechs of all available mechs actually work
2059 => three warnings about OTP not working
2060 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2062 With the patch, sasl_listmech() also gets called three times. But now
2063 SASL's mech_list option is set to the server_mech specified in the the
2064 authenticator. Or in other words, the answer from sasl_listmech()
2065 gets limited to just the mech you're testing for (which is different
2067 => the return list contains just 'plain' or 'login', 'cram-md5' or
2068 nothing depending on the value of ob->server_mech.
2070 I've just tested the patch: Authentication still works fine,
2071 unavailable mechs specified in the exim configuration are still
2072 caught, and the auth.log warnings about OTP are gone."
2074 PH/25 When debugging is enabled, the contents of the command line are added
2075 to the debugging output, even when log_selector=+arguments is not
2078 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2079 answer is "GNU", and only if the return is "GNU/something" is the answer
2082 PH/27 $acl_verify_message is now set immediately after the failure of a
2083 verification in an ACL, and so is available in subsequent modifiers. In
2084 particular, the message can be preserved by coding like this:
2086 warn !verify = sender
2087 set acl_m0 = $acl_verify_message
2089 Previously, $acl_verify_message was set only while expanding "message"
2090 and "log_message" when a very denied access.
2092 PH/28 Modified OS/os.c-Linux with
2094 -#ifndef OS_LOAD_AVERAGE
2095 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2097 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2098 nomenclature these days.)
2100 PH/29 Installed patch from the Sieve maintainer that adds the options
2101 sieve_useraddress and sieve_subaddress to the redirect router.
2103 PH/30 In these circumstances:
2104 . Two addresses routed to the same list of hosts;
2105 . First host does not offer TLS;
2106 . First host accepts first address;
2107 . First host gives temporary error to second address;
2108 . Second host offers TLS and a TLS session is established;
2109 . Second host accepts second address.
2110 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2111 and peerdn, if requested) that were in fact used only for the second
2114 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2115 attention to any local part prefix or suffix that was matched by the
2116 router that accepted the address. It now behaves in the same way as it
2117 does for delivery: the affixes are removed from the local part unless
2118 rcpt_include_affixes is set on the transport.
2120 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2121 timeout during the DATA phase of an incoming message.
2123 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2124 applied a patch sanctioned by the Sieve maintainer.
2126 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2127 the uid or gid is negative. A case of a negative gid caused this to be
2128 noticed. The fix allows for either to be negative.
2130 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2131 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2134 PH/36 The change PH/12 above was broken. Fixed it.
2136 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2137 the grounds that routing the same address twice would always produce the
2138 same answer. This might have been true once, but it is certainly no
2139 longer true now. Routing a child address may depend on the previous
2140 routing that produced that child. Some complicated redirection strategies
2141 went wrong when messages had multiple recipients, and made Exim's
2142 behaviour dependent on the order in which the addresses were given.
2144 I have moved the duplicate checking until after the routing is complete.
2145 Exim scans the addresses that are assigned to local and remote
2146 transports, and removes any duplicates. This means that more work will be
2147 done, as duplicates will always all be routed, but duplicates are
2148 presumably rare, so I don't expect this is of any significance.
2150 For deliveries to pipes, files, and autoreplies, the duplicate checking
2151 still happens during the routing process, since they are not going to be
2154 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2155 It corrects a timeout issue with spamd. This is Ian's comment: "The
2156 background is that sometimes spamd either never reads data from a
2157 connection it has accepted, or it never writes response data. The exiscan
2158 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2159 blindly assumes that writes won't block so it may never time out."
2161 PH/39 Allow G after quota size as well as K and M.
2163 PH/40 The value set for $authenticated_id in an authenticator may not contain
2164 binary zeroes or newlines because the value is written to log lines and
2165 to spool files. There was no check on this. Now the value is run through
2166 the string_printing() function so that such characters are converted to
2167 printable escape sequences.
2169 PH/41 $message_linecount is a new variable that contains the total number of
2170 lines in the message. Compare $body_linecount, which is the count for the
2173 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2174 bounce and delay warning messages, except in certain special cases, which
2177 (a) An SMTP error message from a remote host;
2178 (b) A message specified in a :fail: redirection;
2179 (c) A message specified in a "fail" command in a system filter;
2180 (d) A message specified in a FAIL return from the queryprogram router;
2181 (e) A message specified by the cannot_route_message router option.
2183 In these cases only, Exim does include the error details in bounce and
2184 warning messages. There are also a few cases where bland messages such
2185 as "unrouteable address" or "local delivery error" are given.
2187 PH/43 $value is now also set for the "else" part of a ${run expansion.
2189 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2190 being worked on, but at least Exim now implements the latest version to
2193 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2194 process to complete was treated as a delivery failure, a timeout while
2195 writing the message to the pipe was logged, but erroneously treated as a
2196 successful delivery. Such timeouts include transport filter timeouts. For
2197 consistency with the overall process timeout, these timeouts are now
2198 treated as errors, giving rise to delivery failures by default. However,
2199 there is now a new Boolean option for the pipe transport called
2200 timeout_defer, which, if set TRUE, converts the failures into defers for
2201 both kinds of timeout. A transport filter timeout is now identified in
2204 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2205 systems where "make" and "gmake" are different, calling "gmake" at top
2206 level broke things. I've arranged for the value of $(MAKE) to be passed
2207 from the Makefile to this script so that it can call the same version of
2211 A note about Exim versions 4.44 and 4.50
2212 ----------------------------------------
2214 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2215 changes of various kinds. As a consequence, a big documentation update was
2216 needed. This delayed the release for rather longer than seemed good, especially
2217 in the light of a couple of (minor) security issues. Therefore, the changes
2218 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2219 release. So 4.44 and 4.50 are in effect two different branches that both start
2222 I have left the 4.50 change log unchanged; it contains all the changes since
2223 4.43. The change log for 4.44 is below; many of its items are identical to
2224 those for 4.50. This seems to be the most sensible way to preserve the
2225 historical information.
2231 1. Minor wording change to the doc/README.SIEVE file.
2233 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2234 computation of the current number of files was incorrect.
2236 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2237 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2240 4. Give more explanation in the error message when the command for a transport
2241 filter fails to execute.
2243 5. There are several places where Exim runs a non-Exim command in a
2244 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2245 was being done only for the command run by the queryprogram router. It is
2246 now done for all such subprocesses. The other cases are: ${run, transport
2247 filters, and the commands run by the lmtp and pipe transports.
2249 6. Added CONFIGURE_GROUP build-time option.
2251 7. Some older OS have a limit of 256 on the maximum number of file
2252 descriptors. Exim was using setrlimit() to set 1000 as a large value
2253 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2254 systems. I've change it so that if it can't get 1000, it tries for 256.
2256 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2257 was an oversight, and furthermore, ever since the addition of extra
2258 controls (e.g. 4.43/32), the checks on when to allow different forms of
2259 "control" were broken. There should now be diagnostics for all cases when a
2260 control that does not make sense is encountered.
2262 9. Added the /retain_sender option to "control=submission".
2264 10. $recipients is now available in the predata ACL (oversight).
2266 11. Tidy the search cache before the fork to do a delivery from a message
2267 received from the command line. Otherwise the child will trigger a lookup
2268 failure and thereby defer the delivery if it tries to use (for example) a
2269 cached ldap connection that the parent has called unbind on.
2271 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2272 of $address_data from the recipient verification was clobbered by the
2273 sender verification.
2275 13. The value of address_data from a sender verification is now available in
2276 $sender_address_data in subsequent conditions in the ACL statement.
2278 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2280 15. Added a new option "connect=<time>" to callout options, to set a different
2283 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2284 was its contents. (It was OK if the option was not defined at all.)
2286 17. A "Completed" log line is now written for messages that are removed from
2287 the spool by the -Mrm option.
2289 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2290 information about exactly what failed.
2292 19. Added -dd to debug only the daemon process.
2294 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2295 handles timeouts, both on the server side and network timeouts. Renamed the
2296 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2298 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2299 It is now set to "smtps".
2301 22. $host_address is now set to the target address during the checking of
2302 ignore_target_hosts.
2304 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2305 being passed; this would have caused $sender_host_name to have been used if
2306 matching the list had actually called for a host name (not very likely,
2307 since this list is usually IP addresses). A host name is now passed as
2310 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2311 code that specifies a non-restarting handler (typically sigaction() in
2312 modern systems) in an attempt to fix a rare and obscure crash bug.
2314 25. Narrowed the window for a race in the daemon that could cause it to ignore
2315 SIGCHLD signals. This is not a major problem, because they are used only to
2316 wake it up if nothing else does.
2318 26. A malformed maildirsize file could cause Exim to calculate negative values
2319 for the mailbox size or file count. Odd effects could occur as a result.
2320 The maildirsize information is now recalculated if the size or filecount
2323 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2324 support for a long time. Removed HAVE_SYS_VFS_H.
2326 28. Installed the latest version of exipick from John Jetmore.
2328 29. In an address list, if the pattern was not a regular expression, an empty
2329 subject address (from a bounce message) matched only if the pattern was an
2330 empty string. Non-empty patterns were not even tested. This was the wrong
2331 because it is perfectly reasonable to use an empty address as part of a
2332 database query. An empty address is now tested by patterns that are
2333 lookups. However, all the other forms of pattern expect the subject to
2334 contain a local part and a domain, and therefore, for them, an empty
2335 address still always fails if the pattern is not itself empty.
2337 30. Exim went into a mad DNS loop when attempting to do a callout where the
2338 host was specified on an smtp transport, and looking it up yielded more
2339 than one IP address.
2341 31. Re-factored the code for checking spool and log partition space into a
2342 function that finds that data and another that does the check. The former
2343 is then used to implement four new variables: $spool_space, $log_space,
2344 $spool_inodes, and $log_inodes.
2346 32. The RFC2047 encoding function was originally intended for short strings
2347 such as real names; it was not keeping to the 75-character limit for
2348 encoded words that the RFC imposes. It now respects the limit, and
2349 generates multiple encoded words if necessary. To be on the safe side, I
2350 have increased the buffer size for the ${rfc2047: expansion operator from
2353 33. It is now permitted to omit both strings after an "if" condition; if the
2354 condition is true, the result is "true". As before, when the second string
2355 is omitted, a false condition yields an empty string. This makes it less
2356 cumbersome to write custom ACL and router conditions.
2358 34. Failure to deliver a bounce message always caused it to be frozen, even if
2359 there was an errors_to setting on the router. The errors_to setting is now
2362 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2363 canonical form (fully expanded) before being placed in
2364 $sender_host_address.
2366 36. The table in the code that translates DNS record types into text (T_A to
2367 "A" for instance) was missing entries for NS and CNAME. It is just possible
2368 that this could have caused confusion if both these types were looked up
2369 for the same domain, because the text type is used as part of Exim's
2370 per-process caching. But the chance of anyone hitting this buglet seems
2373 37. The dnsdb lookup has been extended in a number of ways.
2375 (1) There is a new type, "zns", which walks up the domain tree until it
2376 finds some nameserver records. It should be used with care.
2378 (2) There is a new type, "mxh", which is like "mx" except that it returns
2379 just the host names, not the priorities.
2381 (3) It is now possible to give a list of domains (or IP addresses) to be
2382 looked up. The behaviour when one of the lookups defers can be
2383 controlled by a keyword.
2385 (4) It is now possible to specify the separator character for use when
2386 multiple records are returned.
2388 38. The dnslists ACL condition has been extended: it is now possible to supply
2389 a list of IP addresses and/or domains to be looked up in a particular DNS
2392 39. Added log_selector=+queue_time_overall.
2394 40. When running the queue in the test harness, wait just a tad after forking a
2395 delivery process, to get repeatability of debugging output.
2397 41. Include certificate and key file names in error message when GnuTLS fails
2398 to set them up, because the GnuTLS error message doesn't include the name
2399 of the failing file when there is a problem reading it.
2401 42. Allow both -bf and -bF in the same test run.
2403 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2405 44. The "Exiscan patch" is now merged into the mainline Exim source.
2407 45. Sometimes the final signoff response after QUIT could fail to get
2408 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2409 < 0 before doing a fflush(). This bug looks as though it goes back to the
2410 introduction of TLS in release 3.20, but "sometimes" must have been rare
2411 because the tests only now provoked it.
2413 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2414 (this can affect the format of dates).
2416 47. exim_tidydb, when checking for the continued existence of a message for
2417 which it has found a message-specific retry record, was not finding
2418 messages that were in split spool directories. Consequently, it was
2419 deleting retry records that should have stayed in existence.
2421 48. Steve fixed some bugs in eximstats.
2423 49. The SPA authentication driver was not abandoning authentication and moving
2424 on to the next authenticator when an expansion was forced to fail,
2425 contradicting the general specification for all authenticators. Instead it
2426 was generating a temporary error. It now behaves as specified.
2428 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2429 (the order specifies the preference for clients). The order is now AES256,
2430 AES128, 3DES, ARCFOUR128.
2432 51. Small patch to Sieve code - explicitly set From: when generating an
2435 52. Exim crashed if a remote delivery caused a very long error message to be
2436 recorded - for instance if somebody sent an entire SpamAssassin report back
2437 as a large number of 550 error lines. This bug was coincidentally fixed by
2438 increasing the size of one of Exim's internal buffers (big_buffer) that
2439 happened as part of the Exiscan merge. However, to be on the safe side, I
2440 have made the code more robust (and fixed the comments that describe what
2443 53. Now that there can be additional text after "Completed" in log lines (if
2444 the queue_time_overall log selector is set), a one-byte patch to exigrep
2445 was needed to allow it to recognize "Completed" as not the last thing in
2448 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2449 patch that reportedly fixes this has been added. I am not expert enough to
2450 create a test for it. This is what the patch creator wrote:
2452 "I found a little strange behaviour of ldap code when working with
2453 Windows 2003 AD Domain, where users was placed in more than one
2454 Organization Units. When I tried to give exim partial DN, the exit code
2455 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2456 But simultaneously result of request was absolutely normal ldap result,
2457 so I produce this patch..."
2459 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2460 REFERENCE, so I have modified the code to exclude the patch when that macro
2463 55. Some experimental protocols are using DNS PTR records for new purposes. The
2464 keys for these records are domain names, not reversed IP addresses. The
2465 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2466 leaves it alone. Component reversal etc. now happens only for IP addresses.
2469 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2471 57. Double the size of the debug message buffer (to 2048) so that more of very
2472 long debug lines gets shown.
2474 58. The exicyclog utility now does better if the number of log files to keep
2475 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2477 59. Two changes related to the smtp_active_hostname option:
2479 (1) $smtp_active_hostname is now available as a variable.
2480 (2) The default for smtp_banner uses $smtp_active_hostname instead
2481 of $primary_hostname.
2483 60. The host_aton() function is supposed to be passed a string that is known
2484 to be a valid IP address. However, in the case of IPv6 addresses, it was
2485 not checking this. This is a hostage to fortune. Exim now panics and dies
2486 if the condition is not met. A case was found where this could be provoked
2487 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2488 components; fortuitously, this particular loophole had already been fixed
2489 by change 4.50/55 above.
2491 If there are any other similar loopholes, the new check in host_aton()
2492 itself should stop them being exploited. The report I received stated that
2493 data on the command line could provoke the exploit when Exim was running as
2494 exim, but did not say which command line option was involved. All I could
2495 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2496 running as the user.
2499 61. There was a buffer overflow vulnerability in the SPA authentication code
2500 (which came originally from the Samba project). I have added a test to the
2501 spa_base64_to_bits() function which I hope fixes it.
2504 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2505 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2507 63. The daemon start-up calls getloadavg() while still root for those OS that
2508 need the first call to be done as root, but it missed one case: when
2509 deliver_queue_load_max is set with deliver_drop_privilege. This is
2510 necessary for the benefit of the queue runner, because there is no re-exec
2511 when deliver_drop_privilege is set.
2513 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2514 This has been fixed.
2516 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2517 was in use, was not putting the data itself into the right store pool;
2518 consequently, it could be overwritten for a subsequent message in the same
2519 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2522 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2524 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2525 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2526 the function were treating the return as a boolean value, which happened to
2527 work because 0=false and not-0=true, but is not correct code.
2529 68. The host_aton() function was not handling scoped IPv6 addresses (those
2530 with, for example, "%eth0" on the end) correctly.
2532 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2533 negated items (that is, ~something) in unsigned ints. Some compilers
2534 apparently mutter when there is no cast.
2536 70. If an address verification called from an ACL failed, and did not produce a
2537 user-specific message (i.e. there was only a "system" message), nothing was
2538 put in $acl_verify_message. In this situation, it now puts the system
2541 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2542 session; change 4.31/43 added the unwanted input to the log line - except
2543 that it did not do this in the start of session case. It now does.
2545 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2546 This is wrong and can cause the other end to generate a synchronization
2547 error if it is another Exim or anything else that does the synchronization
2548 check. A QUIT command is no longer sent after a timeout.
2550 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2551 during host lookups.
2553 74. The defer_ok option of callout verification was not working if it was used
2554 when verifying addresses in header lines, that is, for this case:
2556 verify = header_sender/callout=defer_ok
2558 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2559 those file descriptors could be used for SMTP connections. If anything
2560 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2561 could be sent to the SMTP client, causing chaos. The daemon now opens
2562 stdin, stdout, and stderr to /dev/null when it puts itself into the
2565 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2566 log by default. The user can override this with suitable Perl magic.
2568 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2569 the log message when discard triggers, was not working for the DATA ACL or
2570 for the non-SMTP ACL.
2572 78. Error message wording change in sieve.c.
2574 79. If smtp_accept_max_per_host was set, the number of connections could be
2575 restricted to fewer than expected, because the daemon was trying to set up
2576 a new connection before checking whether the processes handling previous
2577 connections had finished. The check for completed processes is now done
2578 earlier. On busy systems, this bug wouldn't be noticed because something
2579 else would have woken the daemon, and it would have reaped the completed
2582 80. If a message was submitted locally by a user whose login name contained one
2583 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2584 It caused a spool format error. I have fixed the spool reading code. A
2585 related problem was that the "from" clause in the Received: line became
2586 illegal because of the space(s). It is now covered by ${quote_local_part.
2588 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2591 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2593 83. Patch from Sieve maintainer to fix unterminated string problem in
2594 "vacation" handling.
2596 84. Some minor changes to the Linux configuration files to help with other
2597 OS variants using glibc.
2599 85. One more patch for Sieve to update vacation handling to latest spec.
2602 ----------------------------------------------------
2603 See the note above about the 4.44 and 4.50 releases.
2604 ----------------------------------------------------
2610 1. Change 4.43/35 introduced a bug that caused file counts to be
2611 incorrectly computed when quota_filecount was set in an appendfile
2614 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2615 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2618 3. Give more explanation in the error message when the command for a transport
2619 filter fails to execute.
2621 4. There are several places where Exim runs a non-Exim command in a
2622 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2623 was being done only for the command run by the queryprogram router. It is
2624 now done for all such subprocesses. The other cases are: ${run, transport
2625 filters, and the commands run by the lmtp and pipe transports.
2627 5. Some older OS have a limit of 256 on the maximum number of file
2628 descriptors. Exim was using setrlimit() to set 1000 as a large value
2629 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2630 systems. I've change it so that if it can't get 1000, it tries for 256.
2632 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2633 was an oversight, and furthermore, ever since the addition of extra
2634 controls (e.g. 4.43/32), the checks on when to allow different forms of
2635 "control" were broken. There should now be diagnostics for all cases when a
2636 control that does not make sense is encountered.
2638 7. $recipients is now available in the predata ACL (oversight).
2640 8. Tidy the search cache before the fork to do a delivery from a message
2641 received from the command line. Otherwise the child will trigger a lookup
2642 failure and thereby defer the delivery if it tries to use (for example) a
2643 cached ldap connection that the parent has called unbind on.
2645 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2646 of $address_data from the recipient verification was clobbered by the
2647 sender verification.
2649 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2650 was its contents. (It was OK if the option was not defined at all.)
2652 11. A "Completed" log line is now written for messages that are removed from
2653 the spool by the -Mrm option.
2655 12. $host_address is now set to the target address during the checking of
2656 ignore_target_hosts.
2658 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2659 being passed; this would have caused $sender_host_name to have been used if
2660 matching the list had actually called for a host name (not very likely,
2661 since this list is usually IP addresses). A host name is now passed as
2664 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2665 code that specifies a non-restarting handler (typically sigaction() in
2666 modern systems) in an attempt to fix a rare and obscure crash bug.
2668 15. Narrowed the window for a race in the daemon that could cause it to ignore
2669 SIGCHLD signals. This is not a major problem, because they are used only to
2670 wake it up if nothing else does.
2672 16. A malformed maildirsize file could cause Exim to calculate negative values
2673 for the mailbox size or file count. Odd effects could occur as a result.
2674 The maildirsize information is now recalculated if the size or filecount
2677 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2678 support for a long time. Removed HAVE_SYS_VFS_H.
2680 18. Updated exipick to current release from John Jetmore.
2682 19. Allow an empty sender to be matched against a lookup in an address list.
2683 Previously the only cases considered were a regular expression, or an
2686 20. Exim went into a mad DNS lookup loop when doing a callout where the
2687 host was specified on the transport, if the DNS lookup yielded more than
2690 21. The RFC2047 encoding function was originally intended for short strings
2691 such as real names; it was not keeping to the 75-character limit for
2692 encoded words that the RFC imposes. It now respects the limit, and
2693 generates multiple encoded words if necessary. To be on the safe side, I
2694 have increased the buffer size for the ${rfc2047: expansion operator from
2697 22. Failure to deliver a bounce message always caused it to be frozen, even if
2698 there was an errors_to setting on the router. The errors_to setting is now
2701 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2702 canonical form (fully expanded) before being placed in
2703 $sender_host_address.
2705 24. Updated eximstats to version 1.33
2707 25. Include certificate and key file names in error message when GnuTLS fails
2708 to set them up, because the GnuTLS error message doesn't include the name
2709 of the failing file when there is a problem reading it.
2711 26. Expand error message when OpenSSL has problems setting up cert/key files.
2714 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2715 (this can affect the format of dates).
2717 28. exim_tidydb, when checking for the continued existence of a message for
2718 which it has found a message-specific retry record, was not finding
2719 messages that were in split spool directories. Consequently, it was
2720 deleting retry records that should have stayed in existence.
2722 29. eximstats updated to version 1.35
2723 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2724 1.35 - bugfix such that pie charts by volume are generated correctly
2726 30. The SPA authentication driver was not abandoning authentication and moving
2727 on to the next authenticator when an expansion was forced to fail,
2728 contradicting the general specification for all authenticators. Instead it
2729 was generating a temporary error. It now behaves as specified.
2731 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2732 (the order specifies the preference for clients). The order is now AES256,
2733 AES128, 3DES, ARCFOUR128.
2735 31. Small patch to Sieve code - explicitly set From: when generating an
2738 32. Exim crashed if a remote delivery caused a very long error message to be
2739 recorded - for instance if somebody sent an entire SpamAssassin report back
2740 as a large number of 550 error lines. This bug was coincidentally fixed by
2741 increasing the size of one of Exim's internal buffers (big_buffer) that
2742 happened as part of the Exiscan merge. However, to be on the safe side, I
2743 have made the code more robust (and fixed the comments that describe what
2746 33. Some experimental protocols are using DNS PTR records for new purposes. The
2747 keys for these records are domain names, not reversed IP addresses. The
2748 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2749 leaves it alone. Component reversal etc. now happens only for IP addresses.
2752 34. The host_aton() function is supposed to be passed a string that is known
2753 to be a valid IP address. However, in the case of IPv6 addresses, it was
2754 not checking this. This is a hostage to fortune. Exim now panics and dies
2755 if the condition is not met. A case was found where this could be provoked
2756 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2757 components; fortuitously, this particular loophole had already been fixed
2758 by change 4.50/55 or 4.44/33 above.
2760 If there are any other similar loopholes, the new check in host_aton()
2761 itself should stop them being exploited. The report I received stated that
2762 data on the command line could provoke the exploit when Exim was running as
2763 exim, but did not say which command line option was involved. All I could
2764 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2765 running as the user.
2768 35. There was a buffer overflow vulnerability in the SPA authentication code
2769 (which came originally from the Samba project). I have added a test to the
2770 spa_base64_to_bits() function which I hope fixes it.
2773 36. The daemon start-up calls getloadavg() while still root for those OS that
2774 need the first call to be done as root, but it missed one case: when
2775 deliver_queue_load_max is set with deliver_drop_privilege. This is
2776 necessary for the benefit of the queue runner, because there is no re-exec
2777 when deliver_drop_privilege is set.
2779 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2780 was in use, was not putting the data itself into the right store pool;
2781 consequently, it could be overwritten for a subsequent message in the same
2782 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2785 38. Sometimes the final signoff response after QUIT could fail to get
2786 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2787 < 0 before doing a fflush(). This bug looks as though it goes back to the
2788 introduction of TLS in release 3.20, but "sometimes" must have been rare
2789 because the tests only now provoked it.
2795 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2796 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2797 Now it returns an integer. A number of calls were still expecting a T/F
2798 return. Fortuitously, in all cases, the tests worked in OK situations,
2799 which is the norm. However, things would have gone wrong on any write
2800 failures on the smtp file descriptor. This function is used when sending
2801 messages over SMTP and also when doing verify callouts.
2803 2. When Exim is called to do synchronous delivery of a locally submitted
2804 message (the -odf or -odi options), it no longer closes stderr before doing
2807 3. Implemented the mua_wrapper option.
2809 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2811 5. Implemented the functions header_remove(), header_testname(),
2812 header_add_at_position(), and receive_remove_recipient(), and exported them
2815 6. If an ACL "warn" statement specified the addition of headers, Exim already
2816 inserted X-ACL-Warn: at the start if there was no header name. However, it
2817 was not making this test for the second and subsequent header lines if
2818 there were newlines in the string. This meant that an invalid header could
2819 be inserted if Exim was badly configured.
2821 7. Allow an ACL "warn" statement to add header lines at the start or after all
2822 the Received: headers, as well as at the end.
2824 8. Added the rcpt_4xx retry error code.
2826 9. Added postmaster_mailfrom=xxx to callout verification option.
2828 10. Added mailfrom=xxxx to the callout verification option, for verify=
2831 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2832 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2834 12. Inserted some casts to stop certain compilers warning when using pointer
2835 differences as field lengths or precisions in printf-type calls (mostly
2836 affecting debugging statements).
2838 13. Added optional readline() support for -be (dynamically loaded).
2840 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2841 same clock tick as a message's arrival, so that its received time was the
2842 same as the "first fail" time on the retry record, and that message
2843 remained on the queue past the ultimate address timeout, every queue runner
2844 would try a delivery (because it was past the ultimate address timeout) but
2845 after another failure, the ultimate address timeout, which should have then
2846 bounced the address, did not kick in. This was a "< instead of <=" error;
2847 in most cases the first failure would have been in the next clock tick
2848 after the received time, and all would be well.
2850 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2851 being recognized when the domain list was tested by the match_domain
2852 condition in an expansion string.
2854 16. Added the ${str2b64: operator.
2856 17. Exim was always calling setrlimit() to set a large limit for the number of
2857 processes, without checking whether the existing limit was already
2858 adequate. (It did check for the limit on file descriptors.) Furthermore,
2859 errors from getrlimit() and setrlimit() were being ignored. Now they are
2860 logged to the main and panic logs, but Exim does carry on, to try to do its
2861 job under whatever limits there are.
2863 18. Imported PCRE 5.0.
2865 19. Trivial typo in log message " temporarily refused connection" (the leading
2868 20. If the log selector return_path_on_delivery was set and an address was
2869 redirected to /dev/null, the delivery process crashed because it assumed
2870 that a return path would always be set for a "successful" delivery. In this
2871 case, the whole delivery is bypassed as an optimization, and therefore no
2874 21. Internal re-arrangement: the function for sending a challenge and reading
2875 a response while authentication was assuming a zero-terminated challenge
2876 string. It's now changed to take a pointer and a length, to allow for
2877 binary data in such strings.
2879 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2881 23. Exim was not respecting finduser_retries when seeking the login of the
2882 uid under which it was called; it was always trying 10 times. (The default
2883 setting of finduser_retries is zero.) Also, it was sleeping after the final
2884 failure, which is pointless.
2886 24. Implemented tls_on_connect_ports.
2888 25. Implemented acl_smtp_predata.
2890 26. If the domain in control=submission is set empty, Exim assumes that the
2891 authenticated id is a complete email address when it generates From: or
2892 Sender: header lines.
2894 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2895 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2896 chown and chgrp in /bin and hostname in /usr/bin.
2898 28. Exim was keeping the "process log" file open after each use, just as it
2899 does for the main log. This opens the possibility of it remaining open for
2900 long periods when the USR1 signal hits a daemon. Occasional processlog
2901 errors were reported, that could have been caused by this. Anyway, it seems
2902 much more sensible not to leave this file open at all, so that is what now
2905 29. The long-running daemon process does not normally write to the log once it
2906 has entered its main loop, and it closes the log before doing so. This is
2907 so that log files can straightforwardly be renamed and moved. However,
2908 there are a couple of unusual error situations where the daemon does write
2909 log entries, and I had neglected to close the log afterwards.
2911 30. The text of an SMTP error response that was received during a remote
2912 delivery was being truncated at 512 bytes. This is too short for some of
2913 the long messages that one sometimes sees. I've increased the limit to
2916 31. It is now possible to make retry rules that apply only when a message has a
2917 specific sender, in particular, an empty sender.
2919 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2920 it possible to be selective about when SMTP synchronization is enforced.
2922 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2924 32. Implemented hosts_connection_nolog.
2926 33. Added an ACL for QUIT.
2928 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2931 35. Added mailbox_size and mailbox_filecount to appendfile.
2933 36. Added control = no_multiline_responses to ACLs.
2935 37. There was a bug in the logic of the code that waits for the clock to tick
2936 in the case where the clock went backwards by a substantial amount such
2937 that the microsecond fraction of "now" was more than the microsecond
2938 fraction of "then" (but the whole seconds number was less).
2940 38. Added support for the libradius Radius client library this is found on
2941 FreeBSD (previously only the radiusclient library was supported).
2947 1. When certain lookups returned multiple values in the form name=value, the
2948 quoting of the values was not always being done properly. Specifically:
2949 (a) If the value started with a double quote, but contained no whitespace,
2951 (b) If the value contained whitespace other than a space character (i.e.
2952 tabs or newlines or carriage returns) it was not quoted.
2953 This fix has been applied to the mysql and pgsql lookups by writing a
2954 separate quoting function and calling it from the lookup code. The fix
2955 should probably also be applied to nisplus, ibase and oracle lookups, but
2956 since I cannot test any of those, I have not disturbed their existing code.
2958 2. A hit in the callout cache for a specific address caused a log line with no
2959 reason for rejecting RCPT. Now it says "Previous (cached) callout
2960 verification failure".
2962 3. There was an off-by-one bug in the queryprogram router. An over-long
2963 return line was truncated at 256 instead of 255 characters, thereby
2964 overflowing its buffer with the terminating zero. As well as fixing this, I
2965 have increased the buffer size to 1024 (and made a note to document this).
2967 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2968 when Exim is waiting for an SMTP response from a remote server, Exim
2969 restarts its select() call on the socket, thereby resetting its timeout.
2970 This is not a problem when such interrupts are rare. Somebody set up a cron
2971 job to run exiwhat every 2 minutes, which is less than the normal select()
2972 timeout (5 or 10 minutes). This meant that the select() timeout never
2973 kicked in because it was always reset. I have fixed this by comparing the
2974 time when an interrupt arrives with the time at the start of the first call
2975 to select(). If more time than the timeout has elapsed, the interrupt is
2976 treated as a timeout.
2978 5. Some internal re-factoring in preparation for the addition of Sieve
2979 extensions (by MH). In particular, the "personal" test is moved to a
2980 separate function, and given an option for scanning Cc: and Bcc: (which is
2981 not set for Exim filters).
2983 6. When Exim created an email address using the login of the caller as the
2984 local part (e.g. when creating a From: or Sender: header line), it was not
2985 quoting the local part when it contained special characters such as @.
2987 7. Installed new OpenBSD configuration files.
2989 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2990 try to make them clearer.
2992 9. Callout options, other than the timeout value, were being ignored when
2993 verifying sender addresses in header lines. For example, when using
2995 verify = header_sender/callout=no_cache
2997 the cache was (incorrectly) being used.
2999 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3000 only the Cygwin environment.
3002 11. When return_path_on_delivery was set as a log selector, if different remote
3003 addresses in the same message used different return paths and parallel
3004 remote delivery occurred, the wrong values would sometimes be logged.
3005 (Whenever a remote delivery process finished, the return path value from
3006 the most recently started remote delivery process was logged.)
3008 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3009 lines when AUTH and/or TLS are in use. This is the "received protocol"
3010 field. Exim used to use "asmtp" for authenticated SMTP, without any
3011 indication (in the protocol name) for TLS use. Now it follows the RFC and
3012 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3013 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3014 names appear in log lines as well as in Received: header lines.
3016 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3017 extensions, and comparison tests, and to fix some bugs.
3019 14. Changes to the "personal" filter test:
3021 (1) The test was buggy in that it was just doing the equivalent of
3022 "contains" tests on header lines. For example, if a user's address was
3023 anne@some.where, the "personal" test would incorrectly be true for
3025 To: susanne@some.where
3027 This test is now done by extracting each address from the header in turn,
3028 and checking the entire address. Other tests that are part of "personal"
3029 are now done using regular expressions (for example, to check local parts
3030 of addresses in From: header lines).
3032 (2) The list of non-personal local parts in From: addresses has been
3033 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3034 taken from the Sieve specification recommendations.
3036 (3) If the message contains any header line starting with "List-" it is
3037 treated as non-personal.
3039 (4) The test for "circular" in the Subject: header line has been removed
3040 because it now seems ill-conceived.
3042 15. Minor typos in src/EDITME comments corrected.
3044 16. Installed latest exipick from John Jetmore.
3046 17. If headers_add on a router specified a text string that was too long for
3047 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3048 of string_sprintf() is now avoided.
3050 18. $message_body_size was not set (it was always zero) when running the DATA
3051 ACL and the local_scan() function.
3053 19. For the "mail" command in an Exim filter, no default was being set for
3054 the once_repeat time, causing a random time value to be used if "once" was
3055 specified. (If the value happened to be <= 0, no repeat happened.) The
3056 default is now 0s, meaning "never repeat". The "vacation" command was OK
3057 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3058 (I found it when inspecting the code).
3060 20. There is now an overall timeout for performing a callout verification. It
3061 defaults to 4 times the callout timeout, which applies to individual SMTP
3062 commands during the callout. The overall timeout applies when there is more
3063 than one host that can be tried. The timeout is checked before trying the
3064 next host. This prevents very long delays if there are a large number of
3065 hosts and all are timing out (e.g. when the network connections are timing
3066 out). The value of the overall timeout can be changed by specifying an
3067 additional sub-option for "callout", called "maxwait". For example:
3069 verify = sender/callout=5s,maxwait=20s
3071 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3072 to the end before writing, but this should make it even safer).
3074 22. Exim was forgetting that it had advertised PIPELINING for the second and
3075 subsequent messages on an SMTP connection. It was also not resetting its
3076 memory on STARTTLS and an internal HELO.
3078 23. When Exim logs an SMTP synchronization error within a session, it now
3079 records whether PIPELINING has been advertised or not.
3081 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3082 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3083 rather than long int.
3085 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3087 26. Added the never_mail option to autoreply.
3093 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3094 crash if the getsockname() call failed; this can happen if a connection is
3095 closed very soon after it is established. The problem was simply in the
3096 order in which certain operations were done, causing Exim to try to write
3097 to the SMTP stream before it had set up the file descriptor. The bug has
3098 been fixed by making things happen in the correct order.
3104 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3105 before the connection was closed, thus losing the rejection response.
3107 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3108 some early Solaris releases, but causes trouble in current releases where
3109 socklen_t is defined.
3111 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3114 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3117 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3118 that releasing the top of it at the end releases what was used for sub-
3119 expansions (unless the block got too big). However, discard this block if
3120 the first thing is a variable or header, so that we can use its block when
3121 it is dynamic (useful for very large $message_headers, for example).
3123 6. Lookups now cache *every* query, not just the most recent. A new, separate
3124 store pool is used for this. It can be recovered when all lookup caches are
3125 flushed. Lookups now release memory at the end of their result strings.
3126 This has involved some general refactoring of the lookup sources.
3128 7. Some code has been added to the store_xxx() functions to reduce the amount
3129 of flapping under certain conditions.
3131 8. log_incoming_interface used to affect only the <= reception log lines. Now
3132 it causes the local interface and port to be added to several more SMTP log
3133 lines, for example "SMTP connection from", and rejection lines.
3135 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3137 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3139 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3140 could be overwritten at the end of the current message (or the start of a
3141 new message if it was set in a HELO ACL). The value is now preserved for
3142 the duration of the SMTP connection.
3144 12. If a transport had a headers_rewrite setting, and a matching header line
3145 contained an unqualified address, that address was qualified, even if it
3146 did not match any rewriting rules. The underlying bug was that the values
3147 of the flags that permit the existence of unqualified sender and recipient
3148 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3149 non-local messages, and by -bnq for local messages) were not being
3150 preserved with the message after it was received.
3152 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3153 "next input=" as part of the text comprising the host identity instead of
3154 the correct text. The code was using the same buffer for two different
3155 strings. However, depending on which order the printing function evaluated
3156 its arguments, the bug did not always show up. Under Linux, for example, my
3157 test suite worked just fine.
3159 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3160 doesn't work with some older versions of Perl. It has been changed to "my",
3161 which in any case is probably the better facility to use.
3163 15. A really picky compiler found some instances of statements for creating
3164 error messages that either had too many or two few arguments for the format
3167 16. The size of the buffer for calls to the DNS resolver has been increased
3168 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3169 for addresses that have a lot of PTR records. This alleviates a problem; it
3170 does not fully solve it.
3172 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3173 buffer now truncates the list and logs the incident, which is the same
3174 action as happens when Exim is looking up a host name and its aliases.
3175 Previously in this situation something unpredictable would happen;
3176 sometimes it was "internal error: store_reset failed".
3178 18. If a server dropped the connection unexpectedly when an Exim client was
3179 using GnuTLS and trying to read a response, the client delivery process
3180 crashed while trying to generate an error log message.
3182 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3183 string, for example:
3185 warn message = H1: something\nH2: something
3187 the text was added as a single header line from Exim's point of view
3188 though it ended up OK in the delivered message. However, searching for the
3189 second and subsequent header lines using $h_h2: did not work. This has been
3190 fixed. Similarly, if a system filter added multiple headers in this way,
3191 the routers could not see them.
3193 20. Expanded the error message when iplsearch is called with an invalid key to
3194 suggest using net-iplsearch in a host list.
3196 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3197 ACLs are no longer actually imposed (and a message to that effect is
3200 22. If a "gecos" field in a passwd entry contained escaped characters, in
3201 particular, if it contained a \" sequence, Exim got it wrong when building
3202 a From: or a Sender: header from that name. A second bug also caused
3203 incorrect handling when an unquoted " was present following a character
3204 that needed quoting.
3206 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3207 was not being matched caselessly.
3209 24. Arranged for all hyphens in the exim.8 source to be escaped with
3212 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3213 properly. Recipient callout cache records were still being keyed to include
3214 the sender, even when use_sender was set false. This led to far more
3215 callouts that were necessary. The sender is no longer included in the key
3216 when use_sender is false.
3218 26. Added "control = submission" modifier to ACLs.
3220 27. Added the ${base62d: operator to decode base 62 numbers.
3222 28. dnsdb lookups can now access SRV records.
3224 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3225 the configuration file.
3227 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3228 (-v) mode. This makes the output for a verbose queue run more intelligible.
3230 31. Added a use_postmaster feature to recipient callouts.
3232 32. Added the $body_zerocount variable, containing the number of binary zero
3233 bytes in the message body.
3235 33. The time of last modification of the "new" subdirectory is now used as the
3236 "mailbox time last read" when there is a quota error for a maildir
3239 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3241 35. Added +ignore_unknown as a special item in host lists.
3243 36. Code for decoding IPv6 addresses in host lists is now included, even if
3244 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3245 address was recognized as an IP address, but was then not correctly decoded
3246 into binary, causing unexpected and incorrect effects when compared with
3253 1. Very minor rewording of debugging text in manualroute to say "list of
3254 hosts" instead of "hostlist".
3256 2. If verify=header_syntax was set, and a header line with an unqualified
3257 address (no domain) and a large number of spaces between the end of the
3258 name and the colon was received, the reception process suffered a buffer
3259 overflow, and (when I tested it) crashed. This was caused by some obsolete
3260 code that should have been removed. The fix is to remove it!
3262 3. When running in the test harness, delay a bit after writing a bounce
3263 message to get a bit more predictability in the log output.
3265 4. Added a call to search_tidyup() just before forking a reception process. In
3266 theory, someone could use a lookup in the expansion of smtp_accept_max_
3267 per_host which, without the tidyup, could leave open a database connection.
3269 5. Added the variables $recipient_data and $sender_data which get set from a
3270 lookup success in an ACL "recipients" or "senders" condition, or a router
3271 "senders" option, similar to $domain_data and $local_part_data.
3273 6. Moved the writing of debug_print from before to after the "senders" test
3276 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3277 problems for message scanning, either using a data ACL, or using
3278 local_scan() because the Received: header was not generated till after they
3279 were called (in order to set the time as the time of reception completion).
3280 I have revised the way this works. The header is now generated after the
3281 body is received, but before the ACL or local_scan() are called. After they
3282 are run, the timestamp in the header is updated.
3288 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3289 before starting a queue runner without re-exec. This happened only when
3290 deliver_drop_privilege was set or when the Exim user was set to root. The
3291 effect of the bug was that timeouts during subsequent deliveries caused
3292 crashes instead of being properly handled. The handler is now left at its
3293 default (and expected) setting.
3295 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3296 message, again when deliver_drop_privilege is set or Exim is run as root.
3297 The bug described in (1) was not present in this case, but the tidying up
3298 of the other signals was missing. I have made the two cases consistent.
3300 3. The ignore_target_hosts setting on a manualroute router was being ignored
3301 for hosts that were looked up using the /MX notation.
3303 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3306 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3307 operated on the sender address. After changing the $sender_address to <>
3308 for the sender address verify, Exim was re-instated it as the original
3309 (before rewriting) address, but remembering that it had rewritten it, so it
3310 wasn't rewriting it again. This bug also had the effect of breaking the
3311 sender address verification caching when the sender address was rewritten.
3313 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3314 This has been changed so that if the ip literal address matches
3315 ignore_target_hosts, the router declines.
3317 7. Added expansion conditions match_domain, match_address, and match_local_
3318 part (NOT match_host).
3320 8. The placeholder for the Received: header didn't have a length field set.
3322 9. Added code to Exim itself and to exim_lock to test for a specific race
3323 condition that could lead to file corruption when using MBX delivery. The
3324 issue is with the lockfile that is created in /tmp. If this file is removed
3325 after a process has opened it but before that process has acquired a lock,
3326 there is the potential for a second process to recreate the file and also
3327 acquire a lock. This could lead to two Exim processes writing to the file
3328 at the same time. The added code performs the same test as UW imapd; it
3329 checks after acquiring the lock that its file descriptor still refers to
3330 the same named file.
3332 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3333 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3336 11. Added the smtp_active_hostname option. If used, this will typically be made
3337 to depend on the incoming interface address. Because $interface_address is
3338 not set up until the daemon has forked a reception process, error responses
3339 that can happen earlier (such as "too many connections") no longer contain
3342 12. If an expansion in a condition on a "warn" statement fails because a lookup
3343 defers, the "warn" statement is abandoned, and the next ACL statement is
3344 processed. Previously this caused the whole ACL to be aborted.
3346 13. Added the iplsearch lookup type.
3348 14. Added ident_timeout as a log selector.
3350 15. Added tls_certificate_verified as a log selector.
3352 16. Added a global option tls_require_ciphers (compare the smtp transport
3353 option of the same name). This controls incoming TLS connections.
3355 17. I finally figured out how to make tls_require_ciphers do a similar thing
3356 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3357 before starting the TLS session.
3359 18. Tabs are now shown as \t in -bP output.
3361 19. If the log selector return_path_on_delivery was set, Exim crashed when
3362 bouncing a message because it had too many Received: header lines.
3364 20. If two routers both had headers_remove settings, and the first one included
3365 a superfluous trailing colon, the final name in the first list and the
3366 first name in the second list were incorrectly joined into one item (with a
3367 colon in the middle).
3373 1. Added -C and -D options to the exinext utility, mainly to make it easier
3374 to include in the automated testing, but these could be helpful when
3375 multiple configurations are in use.
3377 2. The exinext utility was not formatting the output nicely when there was
3378 an alternate port involved in the retry record key, nor when there was a
3379 message id as well (for retries that were specific to a specific message
3380 and a specific host). It was also confused by IPv6 addresses, because of
3381 the additional colons they contain. I have fixed the IPv4 problem, and
3382 patched it up to do a reasonable job for IPv6.
3384 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3385 delivery, the log line now contains "pipelined" if PIPELINING was used.
3387 4. An SMTP transport process used to panic and die if the bind() call to set
3388 an explicit outgoing interface failed. This has been changed; it is now
3389 treated in the same way as a connect() failure.
3391 5. A reference to $sender_host_name in the part of a conditional expansion
3392 that was being skipped was still causing a DNS lookup. This no longer
3395 6. The def: expansion condition was not recognizing references to header lines
3396 that used bh_ and bheader_.
3398 7. Added the _cache feature to named lists.
3400 8. The code for checking quota_filecount in the appendfile transport was
3401 allowing one more file than it should have been.
3403 9. For compatibility with Sendmail, the command line option
3411 and sets the incoming protocol and host name (for trusted callers). The
3412 host name and its colon can be omitted when only the protocol is to be set.
3413 Note the Exim already has two private options, -pd and -ps, that refer to
3414 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3415 "s", but I don't think that's a major issue.
3417 10. A number of refactoring changes to the code, none of which should affect
3420 (a) The number of logging options was getting close to filling up the
3421 32-bit word that was used as a bit map. I have split them into two classes:
3422 those that are passed in the argument to log_write(), and those that are
3423 only ever tested independently outside of that function. These are now in
3424 separate 32-bit words, so there is plenty of room for expansion again.
3425 There is no change in the user interface or the logging behaviour.
3427 (b) When building, for example, log lines, the code previously used a
3428 macro that called string_cat() twice, in order to add two strings. This is
3429 not really sufficiently general. Furthermore, there was one instance where
3430 it was actually wrong because one of the argument was used twice, and in
3431 one call a function was used. (As it happened, calling the function twice
3432 did not affect the overall behaviour.) The macro has been replaced by a
3433 function that can join an arbitrary number of extra strings onto a growing
3436 (c) The code for expansion conditions now uses a table and a binary chop
3437 instead of a serial search (which was left over from when there were very
3438 few conditions). Also, it now recognizes conditions like "pam" even when
3439 the relevant support is not compiled in: a suitably worded error message is
3440 given if an attempt is made to use such a condition.
3442 11. Added ${time_interval:xxxxx}.
3444 12. A bug was causing one of the ddress fields not to be passed back correctly
3445 from remote delivery subprocesses. The field in question was not being
3446 subsequently used, so this caused to problems in practice.
3448 13. Added new log selectors queue_time and deliver_time.
3450 14. Might have fixed a bug in maildirsizefile handling that threw up
3451 "unexpected character" debug warnings, and recalculated the data
3452 unnecessarily. In any case, I expanded the warning message to give more
3455 15. Added the message "Restricted characters in address" to the statements in
3456 the default ACL that block characters like @ and % in local parts.
3458 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3459 Three changes have been made:
3461 (a) There was a serious bug; a negative response to MAIL caused the whole
3462 recipient domain to be cached as invalid, thereby blocking all messages
3463 to all local parts at the same domain, from all senders. This bug has
3464 been fixed. The domain is no longer cached after a negative response to
3465 MAIL if the sender used is not empty.
3467 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3470 (c) A new callout option, "use_sender" has been added for people who want
3471 the modified behaviour.
3477 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3480 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3481 indeed breaks things for older releases.
3483 3. Added additional logging to the case where there is a problem reading data
3484 from a filter that is running in a subprocess using a pipe, in order to
3485 try to track down a specific problem.
3487 4. Testing facility fudge: when running in the test harness and attempting
3488 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3489 getting "No route to host". Convert this to a timeout.
3491 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3494 6. Some OS don't have socklen_t but use size_t instead. This affects the
3495 fifth argument of getsockopt() amongst other things. This is now
3496 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3497 can be set for individual OS. I have set it for SunOS5, OSF1, and
3498 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3499 some earlier ones do not.
3501 7. Change 4.30/15 was not doing the test caselessly.
3503 8. The standard form for an IPv6 address literal was being rejected by address
3504 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3505 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3506 this, as well as the form without the "IPv6" on the front (but only when
3507 address literals are enabled, of course).
3509 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3511 10. Exim crashed if a message with an empty sender address specified by -f
3512 encountered a router with an errors_to setting. This could be provoked only
3513 by a command such as
3517 where an empty string was supplied; "<>" did not hit this bug.
3519 11. Installed PCRE release 4.5.
3521 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3522 remained set. It is now erased.
3524 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3525 times from message ids (which are base 36 rather than the normal 62).
3527 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3528 were being counted as actual protocol errors, and logged if the log
3529 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3530 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3531 and DATA following a set of rejected RCPTs do not count as protocol errors.
3532 In other words, Exim assumes they were pipelined, though this may not
3533 actually be the case. Of course, in all cases the client gets an
3534 appropriate error code.
3536 15. If a lookup fails in an ACL condition, a message about the failure may
3537 be available; it is used if testing the ACL cannot continue, because most
3538 such messages specify what the cause of the deferral is. However, some
3539 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3540 that caused an old message to be retained and used if a later statement
3541 caused a defer, replacing the real cause of the deferral.
3543 16. If an IP address had so many PTR records that the DNS lookup buffer
3544 was not large enough to hold them, Exim could crash while trying to process
3545 the truncated data. It now detects and logs this case.
3547 17. Further to 4.21/58, another change has been made: if (and only if) the
3548 first line of a message (the first header line) ends with CRLF, a bare LF
3549 in a subsequent header line has a space inserted after it, so as not to
3550 terminate the header.
3552 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3553 unnecessarily, used atoi() instead of strtol(), and didn't check the
3554 termination when getting file sizes from file names by regex.
3556 19. Completely re-implemented the support for maildirsize files, in the light
3557 of a number of problems with the previous contributed implementation
3558 (4.30/29). In particular:
3560 . If the quota is zero, the maildirsize file is maintained, but no quota is
3563 . If the maildir directory does not exist, it is created before any attempt
3564 to write a maildirsize file.
3566 . The quota value in the file is just a cache; if the quota is changed in
3567 the transport, the new value overrides.
3569 . A regular expression is available for excluding directories from the
3572 20. The autoreply transport checks the characters in options that define the
3573 message's headers; it allows continued headers, but it was checking with
3574 isspace() after an embedded newline instead of explicitly looking for a
3577 21. If all the "regular" hosts to which an address was routed had passed their
3578 expiry times, and had not reached their retry times, the address was
3579 bounced, even if fallback hosts were defined. Now Exim should go on to try
3582 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3583 equivalent code in the SMTP transport. Some hosts send humungous responses
3584 to HELO/EHLO, more than 1024 it seems.
3586 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3587 gives compiler warnings in some environments. I've now done it "properly",
3590 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3591 (because of problems with the built-in one) was declared to return uschar *
3592 instead of char *, causing compiler failure.
3594 25. Fixed a file descriptor leak when processing alias/forward files.
3596 26. Fixed a minor format string issue in dbfn.c.
3598 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3600 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3601 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3604 29. When a sender address is verified, it is cached, to save repeating the test
3605 when there is more than one recipient in a message. However, when the
3606 verification involves a callout, it is possible for different callout
3607 options to be set for different recipients. It is too complicated to keep
3608 track of this in the cache, so now Exim always runs a verification when a
3609 callout is required, relying on the callout cache for the optimization.
3610 The overhead is duplication of the address routing, but this should not be
3613 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3614 to be verified with callout=postmaster, and the main callout worked but the
3615 postmaster check failed, the verification correctly failed. However, if a
3616 subsequent RCPT command asked for sender verification *without* the
3617 postmaster check, incorrect caching caused this verification also to fail,
3620 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3621 it was not caching the DNS options (qualify_single, search_parents) that
3622 were used when the lookup failed. A subsequent lookup with different
3623 options therefore always gave the same answer, though there were cases
3624 where it should not have. (Example: a "domains = !$mx_any" option on a
3625 dnslookup router: the "domains" option is always processed without any
3626 widening, but the router might have qualify_single set.) Now Exim uses the
3627 cached value only when the same options are set.
3629 32. Added John Jetmore's "exipick" utility to the distribution.
3631 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3632 than a timeout (e.g. a certificate is required, and is not provided), an
3633 Exim server now closes the connection immediately. Previously it waited for
3634 the client to close - but if the client is SSL, it seems that they each
3635 wait for each other, leading to a delay before one of them times out.
3637 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3638 maintained 0.8.x compatibility because I don't think many are using it, and
3639 it is clearly obsolete.
3641 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3644 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3645 client certificate was expired. A simple patch fixes this, though I don't
3646 understand the full logic of why the verify callback is called multiple
3649 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3650 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3651 which causes problems with some clients (such as the Certicom SSL Plus
3652 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3653 disables the coutermeasure allowing Eudora to connect."
3655 38. Exim was not checking that a write() to a log file succeeded. This could
3656 lead to Bad Things if a log got too big, in particular if it hit a file
3657 size limit. Exim now panics and dies if it cannot write to a log file, just
3658 as it does if it cannot open a log file.
3660 39. Modified OS/Makefile-Linux so that it now contains
3662 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3664 The two -D definitions ensure that Exim is compiled with large file
3665 support, which makes it possible to handle log files that are bigger than
3668 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3669 instance) a domain was checked against a named list that involved a lookup,
3670 causing $domain_data to be set, then another domain was checked against the
3671 same list, then the first domain was re-checked, the value of $domain_data
3672 after the final check could be wrong. In particular, if the second check
3673 failed, it could be set empty. This bug probably also applied to
3676 41. The strip_trailing_dot option was not being applied to the address given
3677 with the -f command-line option.
3679 42. The code for reading a message's header from the spool was incrementing
3680 $received_count, but never initializing it. This meant that the value was
3681 incorrect (doubled) while delivering a message in the same process in which
3682 it was received. In the most common configuration of Exim, this never
3683 happens - a fresh exec is done - but it can happen when
3684 deliver_drop_privilege is set.
3686 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3687 it now includes up to 150 characters of the unexpected data in the log
3690 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3691 and building data strings. The size of both of these buffers was 10 000
3692 bytes - far larger than anybody would *ever* want, thought I. Needless to
3693 say, somebody hit the limit. I have increased the maximum line length to
3694 20 000 and the maximum data length of concatenated lines to 100 000. I have
3695 also fixed two bugs, because there was no checking on these buffers. Tsk,
3696 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3697 buffer is too small.
3699 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3700 lsearch lookups. Now it does.
3702 46. When parsing a route_list item in a manualroute router, a fixed-length
3703 buffer was used for the list of hosts. I made this 1024 bytes long,
3704 thinking that nobody would ever have a list of hosts that long. Wrong.
3705 Somebody had a whole pile of complicated expansion conditions, and the
3706 string was silently truncated, leading to an expansion error. It turns out
3707 that it is easier to change to an unlimited length (owing to other changes
3708 that have happened since this code was originally written) than to build
3709 structure for giving a limitation error. The length of the item that
3710 expands into the list of hosts is now unlimited.
3712 47. The lsearch lookup could not handle data where the length of text line was
3713 more than 4095 characters. Such lines were truncated, leading to shortened
3714 data being returned. It should now handle lines of any length.
3716 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3717 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3720 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3721 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3722 no longer happens in comment lines. A list of replacements is now placed
3723 at the head of all of the source files, except those whose only change is
3724 to replace PERL_COMMAND in the very first #! line.
3726 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3727 messages on the queue, with a bottom-up merge sort, using code contributed
3728 by Michael Haardt. This should make operations like -bp somewhat faster on
3729 large queues. It won't affect queue runners, except when queue_run_in_order
3732 51. Installed eximstats 1.31 in the distribution.
3734 52. Added support for SRV lookups to the dnslookup router.
3736 53. If an ACL referred to $message_body or $message_body_end, the value was not
3737 reset for any messages that followed in the same SMTP session.
3739 54. The store-handling optimization for building very long strings was not
3740 differentiating between the different store pools. I don't think this
3741 actually made any difference in practice, but I've tidied it.
3743 55. While running the routers to verify a sender address, $sender_address
3744 was still set to the sender address. This is wrong, because when routing to
3745 send a bounce to the sender, it would be empty. Therefore, I have changed
3746 it so that, while verifying a sender address, $sender_address is set to <>.
3747 (There is no change to what happens when verifying a recipient address.)
3749 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3750 target A or AAAA records (if not already returned) without resetting the
3751 qualify_single or search_parents options of the DNS resolver. These are
3752 inappropriate in this case because the targets of MX and SRV records must
3753 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3754 target that, when qualified, matched something in the local domain. These
3755 two options are now turned off when doing these lookups.
3757 57. It seems that at least some releases of Reiserfs (which does not have the
3758 concept of a fixed number of inodes) returns zero and not -1 for the
3759 number of available inodes. This interacted badly with check_spool_inodes,
3760 which assumed that -1 was the "no such thing" setting. What I have done is
3761 to check that the total number of inodes is greater than zero before doing
3762 the test of how many are available.
3764 58. When a "warn" ACL statement has a log_message modifier, the message is
3765 remembered, and not repeated. This is to avoid a lot of repetition when a
3766 message has many recipients that cause the same warning to be written.
3767 Howewer, Exim was preserving the list of already written lines for an
3768 entire SMTP session, which doesn't seem right. The memory is now reset if a
3769 new message is started.
3771 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3772 debugging output unless log_rewrite was also set.
3774 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3775 of (int)(handle) when we know that handle contains (void *)(-1).
3777 61. The Exim daemon panic-logs an error return when it closes the incoming
3778 connection. However "connection reset by peer" seems to be common, and
3779 isn't really an error worthy of noting specially, so that particular error
3782 62. When Exim is trying to find all the local interfaces, it used to panic and
3783 die if the ioctl to get the interface flags failed. However, it seems that
3784 on at least one OS (Solaris 9) it is possible to have an interface that is
3785 included in the list of interfaces, but for which you get a failure error
3786 for this call. This happens when the interface is not "plumbed" into a
3787 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3788 failure of the "get flags" call assumes that the interface is down.
3790 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3791 makes life easier for people who are doing arithmetic on fields extracted
3792 from dates, where you often get leading zeros that should not be
3793 interpreted as octal.
3795 64. Added qualify_domain to the redirect router, to override the global
3798 65. If a pathologically long header line contained very many addresses (the
3799 report of this problem mentioned 10 000) and each of them was rewritten,
3800 Exim could use up a very large amount of memory. (It kept on making new
3801 copies of the header line as it rewrote, and never released the old ones.)
3802 At the expense of a bit more processing, the header rewriting function has
3803 been changed so that it no longer eats memory in this way.
3805 66. The generation of the Received: header has been moved from the time that a
3806 message starts to be received, to the time that it finishes. The timestamp
3807 in the Received: header should now be very close to that of the <= log
3808 line. There are two side-effects of this change:
3810 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3811 logged header lines no longer include the local Received: line, because
3812 it has not yet been created. The same applies to a copy of the message
3813 that is returned to a non-SMTP sender when a message is rejected.
3815 (b) When a filter file is tested using -bf, no additional Received: header
3816 is added to the test message. After some thought, I decided that this
3819 This change does not affect the value of $received_for. It is still set
3820 after address rewriting, but before local_scan() is called.
3822 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3824 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3825 gave an unhelpful panic error message, and a defer error. I have managed to
3826 change this behaviour so that it now rejects any supplied certificate,
3827 which seems right, as the list of acceptable certificates is empty.
3829 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3830 gave an unhelpful defer error. I have not managed to make this reject any
3831 supplied certificates, but the error message it gives is "no certificate
3832 supplied", which is not helpful.
3834 70. exigrep's output now also includes lines that are not associated with any
3835 message, but which match the given pattern. Implemented by a patch from
3836 Martin Sluka, which also tidied up the Perl a bit.
3838 71. Recipient callout verification, like sender verification, was using <> in
3839 the MAIL FROM command. This isn't really the right thing, since the actual
3840 sender may affect whether the remote host accepts the recipient or not. I
3841 have changed it to use the actual sender in the callout; this means that
3842 the cache record is now keyed on a recipient/sender pair, not just the
3843 recipient address. There doesn't seem to be a real danger of callout loops,
3844 since a callout by the remote host to check the sender would use <>.
3845 [SEE ABOVE: changed after hitting problems.]
3847 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3848 temporary errors. However, in the case of such a code being given after
3849 the end of a data transmission (i.e. after ".") Exim was failing to write
3850 a retry record for the message. (Yes, there was some broken host that was
3851 actually sending 8xx at this point.)
3853 73. An unknown lookup type in a host list could cause Exim to panic-die when
3854 the list was checked. (An example that provoked this was putting <; in the
3855 middle of a list instead of at the start.) If this happened during a DATA
3856 ACL check, a -D file could be left lying around. This kind of configuration
3857 error no longer causes Exim to die; instead it causes a defer errror. The
3858 incident is still logged to the main and panic logs.
3860 74. Buglet left over from Exim 3 conversion. The message "too many messages
3861 in one connection" was written to the rejectlog but not the mainlog, except
3862 when address rewriting (yes!) was being logged.
3864 75. Added write_rejectlog option.
3866 76. When a system filter was run not as root (that is, when system_filter_user
3867 was set), the values of the $n variables were not being returned to the
3868 main process; thus, they were not subsequently available in the $sn
3871 77. Added +return_path_on_delivery log selector.
3873 78. A connection timeout was being treated differently from recipients deferred
3874 when testing hosts_max_try with a message that was older than the host's
3875 retry timeout. (The host should not be counted, thus allowing all hosts to
3876 be tried at least once before bouncing.) This may have been the cause of an
3877 occasionally reported bug whereby a message would remain on the queue
3878 longer than the retry timeout, but would be bounced if a delivery was
3879 forced. I say "may" because I never totally pinned down the problem;
3880 setting up timeout/retry tests is difficult. See also the next item.
3882 79. The ultimate address timeout was not being applied to errors that involved
3883 a combination of host plus message (for example, a timeout on a MAIL
3884 command). When an address resolved to a number of possible hosts, and they
3885 were not all tried for each delivery (e.g. because of hosts_max_try), a
3886 message could remain on the queue longer than the retry timeout.
3888 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3891 81. Fixed an obscure SMTP outgoing bug which required at least the following
3892 conditions: (a) there was another message waiting for the same server;
3893 (b) the server returned 5xx to all RCPT commands in the first message so
3894 that the message was not completed; (c) the server dropped the connection
3895 or gave a negative response to the RSET that Exim sends to abort the
3896 transaction. The observed case was a dropped connection after DATA that had
3897 been sent in pipelining mode. That is, the server had advertised PIPELINING
3898 but was not implementing it correctly. The effect of the bug was incorrect
3899 behaviour, such as trying another host, and this could lead to a crash.
3905 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3906 and daemon.c were passed as pointers to ints; they should have been
3907 pointers to socklen_t variables (which are typically unsigned ints).
3909 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3912 3. Fixed a really odd bug that affected only the testing scheme; patching a
3913 certain fixed string in the binary changed the value of another string that
3914 happened to be identical to the end of the original first string.
3916 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3917 name", it returns that address as the IP address. On some operating
3918 systems (e.g. Solaris), it also passes back the IP address string as the
3919 "host name". However, on others (e.g. Linux), it passes back an empty
3920 string. Exim wasn't checking for this, and was changing the host name to an
3921 empty string, assuming it had been canonicized.
3923 5. Although rare, it is permitted to have more than one PTR record for a given
3924 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3925 all the names associated with an address, because they do in Solaris.
3926 However, it seems that they do not in Linux for data that comes from the
3927 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3928 I found this out when I moved to a new Linux workstation and tried to run
3929 the Exim test suite.
3931 To get round this problem I have changed the code so that it now does its
3932 own call to the DNS to look up PTR records when searching for a host name.
3933 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3934 addresses that are only in /etc/hosts are still found.
3936 This behaviour is, however, controlled by an option called host_lookup_
3937 order, which defaults to "bydns:byaddr". If people want to use the other
3938 order, or indeed, just use one or the other means of lookup, they can
3939 specify it in this variable.
3941 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3942 some operating systems, this comes back from gethostbyaddr() as an empty
3943 string, and this is what Exim used to test for. However, it seems that in
3944 other systems, "." is yielded. Exim now tests for this case too.
3946 7. The values of check_spool_space and check_log_space are now held internally
3947 as a number of kilobytes instead of an absolute number of bytes. If a
3948 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3949 kilobyte. This means that much larger values can be stored.
3951 8. Exim monitor: an attempt to get the action menu when not actually pointing
3952 at a message produces an empty menu entitled "No message selected". This
3953 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3954 no entries in it ("Shell widget menu has zero width and/or height"). So I
3955 have added a single, blank menu entry in this case.
3957 9. Added ${quote_local_part.
3959 10. MIME decoding is now applied to the contents of Subject: header lines when
3962 11. Now that a reference to $sender_host_address automatically causes a reverse
3963 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3964 host lookup before query-style lookups in lists that might use this
3965 variable. This has therefore been abolished, and the "net-" prefix is no
3966 longer necessary for query-style lookups.
3968 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3969 have been a typo for LFLAGS, so it has been changed.
3971 13. The install script calls Exim with "-C /dev/null" in order to find the
3972 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3973 to be output. Howeve, since Exim outputs its version number before the
3974 error, it didn't break the script. It just looked ugly. I fixed this by
3975 always allowing "-C /dev/null" if the caller is root.
3977 14. Ignore overlarge ACL variable number when reading spool file - insurance
3978 against a later release with more variables having written the file.
3980 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3981 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3982 this, as well as the form without the "IPv6" on the front.
3984 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3985 OS/Makefile-Darwin file.
3987 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3988 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3990 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3991 variable. However, some OS use ints for inodes. I've added cast to long int
3992 to get rid of the compiler warning.
3994 19. I had forgotten to lock out "/../" in configuration file names when
3995 ALT_CONFIG_PREFIX was set.
3997 20. Routers used for verification do not need to specify transports. However,
3998 if such a router generated a host list, and callout was configured, Exim
3999 crashed, because it could not find a port number from the (non-existent)
4000 transport. It now assumes port 25 in this circumstance.
4002 21. Added the -t option to exigrep.
4004 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4005 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4006 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4007 them is set, LOOKUP_LSEARCH is forced.
4009 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4011 24. Added sender and host information to the "rejected by local_scan()" log
4012 line; previously there was no indication of these.
4014 25. Added .include_if_exists.
4016 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4017 Linux. It turns out that not all file systems support this. Apparently some
4018 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4019 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4020 not supported on the file descriptor, is now ignored when Exim is trying to
4021 sync a directory. This applies only to Linux.
4023 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4025 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4026 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4029 29. Added support for maildirsize files from supplied patch (modified a bit).
4031 30. The use of :fail: followed by an empty string could lead Exim to respond to
4032 sender verification failures with (e.g.):
4034 550 Verification failed for <xxx>
4035 550 Sender verify failed
4037 where the first response line was missing the '-' that indicates it is not
4038 the final line of the response.
4040 31. The loop for finding the name of the user that called Exim had a hardwired
4041 limit of 10; it now uses the value of finduser_retries, which is used for
4042 all other user lookups.
4044 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4047 33. Exim was neglecting to zero errno before one call of strtol() when
4048 expanding a string and expecting an integer value. On some systems this
4049 resulted in spurious "integer overflow" errors. Also, it was casting the
4050 result into an int without checking.
4052 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4053 did not work. The code looks as if it has *never* worked, though it appears
4054 to have been documented since at least releast 1.62. I have made it work.
4056 35. The "timeout_DNS" error in retry rules, also documented since at least
4057 1.62, also never worked. As it isn't clear exactly what this means, and
4058 clearly it isn't a major issue, I have abolished the feature by treating it
4059 as "timeout", and writing a warning to the main and panic logs.
4061 36. The display of retry rules for -brt wasn't always showing the error code
4064 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4065 timeout_connect_A, timeout_connect_MX.
4067 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4068 to the empty sender.
4070 39. The daemon was not analysing the content of -oX till after it had closed
4071 stderr and disconnected from the controlling terminal. This meant that any
4072 syntax errors were only noted on the panic log, and the return code from
4073 the command was 0. By re-arranging the code a little, I've made the
4074 decoding happen first, so such errors now appear on stderr, and the return
4075 code is 1. However, the actual setting up of the sockets still happens in
4076 the disconnected process, so errors there are still only recorded on the
4079 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4080 connections (as happens on some IP stacks) was logged at start up time as
4081 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4082 it from "IPv6 and IPv4", which means that two separate sockets are being
4085 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4086 says whether AF_INET or AF_INET6 was passed as an argument.
4088 42. Exiwhat output was messed up when time zones were included in log
4091 43. Exiwhat now gives more information about the daemon's listening ports,
4092 and whether -tls-on-connect was used.
4094 44. The "port" option of the smtp transport is now expanded.
4096 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4097 silently ignored. Now an error message is written to the main and panic
4100 46. There's a new ACL modifier called "logwrite" which writes to a log file
4101 as soon as it is encountered.
4103 47. Added $local_user_uid and $local_user_gid at routing time.
4105 48. Exim crashed when trying to verify a sender address that was being
4108 49. Exim was recognizing only a space character after ".include". It now also
4109 recognizes a tab character.
4111 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4112 extracting the relevant information from the specification. The man page no
4113 longer contains scrambled data for the -d option, and I've added a section
4114 at the front about calling Exim under different names.
4116 51. Added "extra_headers" argument to the "mail" command in filter files.
4118 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4121 53. Installed eximstats 1.29.
4123 54. Added transport_filter_timeout as a generic transport option.
4125 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4126 Cc: header lines. This was required by RFC 822, but it not required by RFC
4129 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4130 incoming messages that did not have them. Now it does so only if the
4131 message originates locally, that is, if there is no associated remote host
4132 address. When Resent- header lines are present, this applies to the Resent-
4133 lines rather than the non-Resent- lines.
4135 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4136 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4138 58. Messages for configuration errors now include the name of the main
4139 configuration file - useful now that there may be more than one file in a
4140 list (.included file names were always shown).
4142 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4143 for those rare installations that do not start the daemon as root or run it
4144 setuid root. I've cut out the call to initgroups() if the daemon is not
4147 60. The Exim user and group can now be bound into the binary as text strings
4148 that are looked up at the start of Exim's processing.
4150 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4152 62. Added $mailstore_basename variable.
4154 63. Installed patch to sieve.c from Michael Haardt.
4156 64. When Exim failed to open the panic log after failing to open the main log,
4157 the original message it was trying to log was written to stderr and debug
4158 output, but if they were not available (the usual case in production), it
4159 was lost. Now it is written to syslog before the two lines that record the
4160 failures to open the logs.
4162 65. Users' Exim filters run in subprocesses under the user's uid. It is
4163 possible for a "deliver" command or an alias in a "personal" command to
4164 provoke an address rewrite. If logging of address rewriting is configured,
4165 this fails because the process is not running as root or exim. There may be
4166 a better way of dealing with this, but for the moment (because 4.30 needs
4167 to be released), I have disabled address rewrite logging when running a
4168 filter in a non-root, non-exim process.
4174 1. The buildconfig auxiliary program wasn't quoting the value set for
4175 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4176 not defined. This bug was masked in 4.22 by the effect that was fixed in
4179 2. Some messages that were rejected after a message id was allocated were
4180 shown as "incomplete" by exigrep. It no longer does this for messages that
4181 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4183 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4184 have allow_domain_literals set, the ID did not get logged in the <= line.
4185 Domain literals are now always recognized in Message-ID: header lines.
4187 4. The first argument for a ${extract expansion item is the key name or field
4188 number. Leading and trailing spaces in this item were not being ignored,
4189 causing some misleading effects.
4191 5. When deliver_drop_privilege was set, single queue runner processes started
4192 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4193 same command in the process it spins off) were not dropping privilege.
4195 6. When the daemon running as "exim" started a queue runner, it always
4196 re-executed Exim in the spun-off process. This is a waste of effort when
4197 deliver_drop_privilege is set. The new process now just calls the
4198 queue-runner function directly.
4204 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4207 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4208 ignored. Though the use of -oP was forcing the writing of a pid file, it
4209 was always written to the default place.
4211 3. If the message "no IP address found for host xxxx" is generated during
4212 incoming verification, it is now followed by identification of the incoming
4213 connection (so you can more easily find what provoked it).
4215 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4217 5. Added some features to "harden" Exim a bit more against certain attacks:
4219 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4220 be put in Local/Makefile. This is like the never_users runtime option,
4221 but it cannot be overridden. The default setting is "root".
4223 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4224 prefix string with which any file named in a -C command line option
4227 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4228 is retained for -C and -D only if the caller of Exim is root. Without
4229 it, the exim user may also use -C and -D and retain privilege.
4231 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4232 command line option is disabled.
4234 6. Macro names set by the -D option must start with an upper case letter, just
4235 like macro names defined in the configuration file.
4237 7. Added "dereference=" facility to LDAP.
4239 8. Two instances of the typo "uknown" in the source files are fixed.
4241 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4242 the Configure-Makefile script screwed up while processing it.
4244 10. Incorporated PCRE 4.4.
4246 11. The SMTP synchronization check was not operating right at the start of an
4247 SMTP session. For example, it could not catch a HELO sent before the client
4248 waited for the greeting. There is now a check for outstanding input at the
4249 point when the greeting is written. Because of the duplex, asynchronous
4250 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4251 way, but not yet received, when the check is performed.
4253 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4254 on TCP/IP sockets, because this apparently causes some broken clients to
4257 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4258 unchanged) from the Cygwin maintainer.
4260 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4261 was supported instead of testing for mbx. Effectively a typo.
4263 15. The spa authenticator server code was not checking that the input it
4264 received was valid base64.
4266 16. The debug output line for the "set" modifier in ACLs was not showing the
4267 name of the variable that was being set.
4269 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4271 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4272 lookup on its own. Something else was needed to trigger the lookup. For
4273 example, a match in host_lookup or the need for a host name in a host list.
4274 Now, if $sender_host_name is referenced and the host name has not yet been
4275 looked up, a lookup is performed. If the lookup fails, the variable remains
4276 empty, and $host_lookup_failed is set to "1".
4278 19. Added "eqi" as a case-independent comparison operator.
4280 20. The saslauthd authentication condition could segfault if neither service
4281 nor realm was specified.
4283 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4284 error message that was logged was misleading, and incoming SMTP
4285 connections were dropped. The message is now more accurate, and temporary
4286 errors are given to SMTP connections.
4288 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4289 (except RSET) with a 421 error, until QUIT is received. However, it was
4290 failing to send a response to QUIT.
4292 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4293 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4294 ACL is now run after the helo_try_verify_hosts code.
4296 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4297 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4298 case-independent, so other case variants are also recognized). Apparently
4299 some systems use these upper case variants.
4301 25. If more than two messages were waiting for the same host, and a transport
4302 filter was specified for the transport, Exim sent two messages over the
4303 same TCP/IP connection, and then failed with "socket operation on non-
4304 socket" when it tried to send the third.
4306 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4308 27. The extern definition of crypt16() in expand.c was not being excluded when
4309 the OS had its own crypt16() function.
4311 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4312 as a preferred synonym for return_size_limit, both as an option and as an
4315 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4317 30. Changed the default configuration ACL to relax the local part checking rule
4318 for addresses that are not in any local domains. For these addresses,
4319 slashes and pipe symbols are allowed within local parts, but the sequence
4320 /../ is explicitly forbidden.
4322 31. SPA server authentication was not clearing the challenge buffer before
4325 32. log_message in a "warn" ACL statement was writing to the reject log as
4326 well as to the main log, which contradicts the documentation and doesn't
4327 seem right (because no rejection is happening). So I have stopped it.
4329 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4330 However, I am unable to do any testing of this.
4332 34. Fixed an infelicity in the appendfile transport. When checking directories
4333 for a mailbox, to see if any needed to be created, it was accidentally
4334 using path names with one or more superfluous leading slashes; tracing
4335 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4337 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4338 contents are added to the log line that is written for every discarded
4339 recipient. (Previously a log_message setting was ignored.)
4341 36. The ${quote: operator now quotes the string if it is empty.
4343 37. The install script runs exim in order to find its version number. If for
4344 some reason other than non-existence or emptiness, which it checks, it
4345 could not run './exim', it was installing it with an empty version number,
4346 i.e. as "exim-". This error state is now caught, and the installation is
4349 38. An argument was missing from the function that creates an error message
4350 when Exim fails to connect to the socket for saslauthd authentication.
4351 This could cause Exim to crash, or give a corrupted message.
4353 39. Added isip, isip4, and isip6 to ${if conditions.
4355 40. The ACL variables $acl_xx are now saved with the message, and can be
4356 accessed later in routers, transports, and filters.
4358 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4359 strings in the file are not string-expanded.
4361 42. If a MAIL command specified a SIZE value that was too large to fit into an
4362 int variable, the check against message_size_limit failed. Such values are
4363 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4364 day this will have to be increased, but I don't think I want to be around
4365 when emails are that large.
4372 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4373 iconv() is not standard in FreeBSD.
4375 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4376 IPv6 enabled. The observed symptom was a segmentation fault on return from
4377 the function os_common_find_running_interfaces() in src/os.c.
4379 3. In the check_special_case() function in daemon.c I had used "errno" as an
4380 argument name, which causes warnings on some systems. This was basically a
4381 typo, since it was named "eno" in the comments!
4383 4. The code that waits for the clock to tick (at a resolution of some fraction
4384 of a second) so as to ensure message-id uniqueness was always waiting for
4385 at least one whole tick, when it could have waited for less. [This is
4386 almost certainly not relevant at current processor speeds, where it is
4387 unlikely to ever wait at all. But we try to future-proof.]
4389 5. The function that sleeps for a time interval that includes fractions of a
4390 second contained a race. It did not block SIGALRM between setting the
4391 timer, and suspending (a couple of lines later). If the interval was short
4392 and the sigsuspend() was delayed until after it had expired, the suspension
4393 never ended. On busy systems this could lead to processes getting stuck for
4396 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4397 process, before it forks any delivery processes. The open lookup caching
4398 mechanism meant that the open file or database connection was passed into
4399 the delivery process. The problem was that delivery processes always tidy
4400 up cached lookup data. This could cause a problem for the next delivery
4401 process started by the queue runner, because the external queue runner
4402 process does not know about the closure. So the next delivery process
4403 still has data in the lookup cache. In the case of a file lookup, there was
4404 no problem because closing a file descriptor in a subprocess doesn't affect
4405 the parent. However, if the lookup was caching a connection to a database,
4406 the connection was closed, and the second delivery process was likely to
4407 see errors such as "PGSQL: query failed: server closed the connection
4408 unexpectedly". The problem has been fixed by closing all cached lookups
4409 in a queue runner before running a delivery process.
4411 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4412 seem to have the "const" qualifier which it has on other OS. I've
4415 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4416 *of the same type* (client or server) with the same public name that an
4417 error should be diagnosed.
4419 9. When Exim looked up a host name for an IP address, but failed to find the
4420 original IP address when looking up the host name (a safety check), it
4421 output the message "<ip address> does not match any IP for NULL", which was
4422 confusing, to say the least. The bug was that the host name should have
4423 appeared instead of "NULL".
4425 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4426 user that is built into the binary, and the -C or -D options is used, root
4427 privilege is dropped before the configuration file is read. In addition,
4428 logging is switched to stderr instead of the normal log files. If the
4429 configuration then re-defines the Exim user, the unprivileged environment
4430 is probably not what is expected, so Exim logs a panic warning message (but
4433 However, if deliver_drop_privilege is set, the unprivileged state may well
4434 be exactly what is intended, so the warning has been cut out in that case,
4435 and Exim is allowed to try to write to its normal log files.
4441 1. smtp_return_error_details was not giving details for temporary sender
4442 or receiver verification errors.
4444 2. Diagnose a configuration error if two authenticators have the same public
4447 3. Exim used not to create the message log file for a message until the first
4448 delivery attempt. This could be confusing when incoming messages were held
4449 for policy or load reasons. The message log file is now created at the time
4450 the message is received, and an initial "Received" line is written to it.
4452 4. The automatically generated man page for command line options had a minor
4453 bug that caused no ill effects; however, a more serious problem was that
4454 the procedure for building the man page automatically didn't always
4455 operate. Consequently, release 4.20 contains an out-of-date version. This
4456 shouldn't happen again.
4458 5. When building Exim with embedded Perl support, the script that builds the
4459 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4460 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4462 6. The freeze_tell option was not being used for messages that were frozen on
4463 arrival, either by an ACL or by local_scan().
4465 7. Added the smtp_incomplete_transaction log selector.
4467 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4468 was accepting AUTH without a new EHLO.
4470 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4471 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4474 10. Logging of TCP/IP connections (when configured) now happens in the main
4475 daemon process instead of the child process, so that the TCP/IP connection
4476 count is more accurate (but it can never be perfect).
4478 11. The use of "drop" in a nested ACL was not being handled correctly in the
4479 outer ACL. Now, if condition failure induced by the nested "drop" causes
4480 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4481 or "require"), the connection is dropped.
4483 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4484 that yield "discard" can now be used with an "accept" or a "discard" verb,
4485 but an error is generated for any others (because I can't see a useful way
4486 to define what should happen).
4488 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4489 are now processed in the same way as lines in the Exim configuration file.
4490 In particular, continuation lines are supported.
4492 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4494 15. Added -ti meaning -t -i.
4496 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4497 domains, and warn by logging if others are found.
4499 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4500 returned from GIFCONF when Exim is trying to find the list of interfaces on
4501 a host. The code in os.c has been modified to copy each ifreq to an aligned
4502 structure in all cases.
4504 Also, in some cases, the returned ifreq's were being copied to a 'struct
4505 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4506 means the last couple of bytes of an IPv6 address could be chopped if the
4507 ifreq contained only a normal sockaddr (14 bytes storage).
4509 18. Named domain lists were not supported in the hosts_treat_as_local option.
4510 An entry such as +xxxx was not recognized, and was treated as a literal
4513 19. Ensure that header lines added by a DATA ACL are included in the reject log
4514 if the ACL subsequently rejects the message.
4516 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4517 MD5 (which is deprecated).
4519 21. When testing a filter file using -bf, Exim was writing a message when it
4520 took the sender from a "From " line in the message, but it was not doing so
4521 when it took $return_path from a Return-Path: header line. It now does.
4523 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4524 with a valid header line field name (a series of printing characters
4525 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4527 23. Changed "disc" in the source to "disk" to conform to the documentation and
4528 the book and for uniformity.
4530 24. Ignore Sendmail's -Ooption=value command line item.
4532 25. When execve() failed while trying to run a command in a pipe transport,
4533 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4534 could be confused with a return value of 69 from the command itself. This
4535 has been changed to 127, the value the shell returns if it is asked to run
4536 a non-existent command. The wording for the related log line suggests a
4537 non-existent command as the problem.
4539 26. If received_header_text expands to an empty string, do not add a Received:
4540 header line to the message. (Well, it adds a token one on the spool, but
4541 marks it "old" so that it doesn't get used or transmitted.)
4543 27. Installed eximstats 1.28 (addition of -nt option).
4545 28. There was no check for failure on the call to getsockname() in the daemon
4546 code. This can fail if there is a shortage of resources on the system, with
4547 ENOMEM, for example. A temporary error is now given on failure.
4549 29. Contrary to the C standard, it seems that in some environments, the
4550 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4551 program. Exim now does this explicitly; it affects the formatting of
4552 timestamps using strftime().
4554 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4555 complaints. I've now initialized all the variables, to avoid this.
4557 32. Header lines added by a system filter were not being "seen" during
4558 transport-time rewrites.
4560 33. The info_callback() function passed to OpenSSL is set up with type void
4561 (*)(SSL *, int, int), as described somewhere. However, when calling the
4562 function (actually a macro) that sets it up, the type void(*)() is
4563 expected. I've put in a cast to prevent warnings from picky compilers.
4565 34. If a DNS black list lookup found a CNAME record, but there were no A
4566 records associated with the domain it pointed at, Exim crashed.
4568 35. If a DNS black list lookup returned more than one A record, Exim ignored
4569 all but the first. It now scans all returned addresses if a particular IP
4570 value is being sought. In this situation, the contents of the
4571 $dnslist_value variable are a list of all the addresses, separated by a
4574 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4575 to accept a host name and all its aliases if the forward lookup for any of
4576 them yielded the IP address of the incoming connection. Now it accepts only
4577 those names whose forward lookup yields the correct IP address. Any other
4578 names are discarded. This closes a loophole whereby a rogue DNS
4579 administrator could create reverse DNS records to break through a
4580 wildcarded host restriction in an ACL.
4582 37. If a user filter or a system filter that ran in a subprocess used any of
4583 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4584 the wrong values were passed to the pipe command ($thisaddress had the
4585 value of $0, $0 had the value of $1, etc). This bug was introduced by
4586 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4588 38. Improved the line breaking for long SMTP error messages from ACLs.
4589 Previously, if there was no break point between 40 and 75 characters, Exim
4590 left the rest of the message alone. Two changes have been made: (a) I've
4591 reduced the minimum length to 35 characters; (b) if it can't find a break
4592 point between 35 and 75 characters, it looks ahead and uses the first one
4593 that it finds. This may give the occasional overlong line, but at least the
4594 remaining text gets split now.
4596 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4597 file descriptors might be low, and that setting 1000 would always raise it.
4598 It turns out that in some environments, the limit is already over 1000 and
4599 that lowering it causes trouble. So now Exim takes care not to decrease it.
4601 40. When delivering a message, the value of $return_path is set to $sender_
4602 address at the start of routing (routers may change the value). By an
4603 oversight, this default was not being set up when an address was tested by
4604 -bt or -bv, which affected the outcome if any router or filter referred to
4607 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4608 log only when "message" or "log_message" are set. However, if one of the
4609 conditions was an address verification, or a call to a nested ACL, the
4610 messages generated by the underlying test were being passed through. This
4611 no longer happens. The underlying message is available in $acl_verify_
4612 message for both "message" and "log_message" expansions, so it can be
4613 passed through if needed.
4615 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4616 new expansion $bh_ to give the encoded byte string without charset
4617 translation. Translation happens only if iconv() is available; HAVE_ICONV
4618 indicates this at build time. HEADERS_CHARSET gives the charset to
4619 translate to; headers_charset can change it in the configuration, and
4620 "headers charset" can change it in an individual filter file.
4622 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4623 that creates RFC 2047 encoded "words" labels them as that charset instead
4624 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4625 expansion operator; (ii) when Exim creates a From: line for a local
4626 message; (iii) when a header line is rewritten to include a "phrase" part.
4628 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4629 buggy, causing it to skip the first lines of messages whose message ID
4630 ended in 'D'. This would not have bitten before Exim release 4.14, because
4631 message IDs were unlikely to end in 'D' before then. The effect was to have
4632 incorrect size information for certain domains.
4634 45. #include "config.h" was missing at the start of the crypt16.c module. This
4635 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4638 46. If there was a timeout during a "random" callout check, Exim treated it as
4639 a failure of the random address, and carried on sending RSET and the real
4640 address. If the delay was just some slowness somewhere, the response to the
4641 original RCPT would be taken as a response to RSET and so on, causing
4642 mayhem of various kinds.
4644 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4645 when I implemented it. It didn't allow for the fact that some option values
4646 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4647 the right test for positive values.
4649 48. Domain names in DNS records are case-independent. Exim always looks them up
4650 in lower case. Some resolvers return domain names in exactly the case they
4651 appear in the zone file, that is, they may contain uppercase letters. Not
4652 all resolvers do this - some return always lower case. Exim was treating a
4653 change of case by a resolver as a change of domain, similar to a widening
4654 of a domain abbreviation. This triggered its re-routing code and so it was
4655 trying to route what was effectively the same domain again. This normally
4656 caused routing to fail (because the router wouldn't handle the domain
4657 twice). Now Exim checks for this case specially, and just changes the
4658 casing of the domain that it ultimately uses when it transmits the message
4661 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4664 50. If a filter generated a file delivery with a non-absolute name (possible if
4665 no home directory exists for the router), the forbid_file option was not
4668 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4669 the existing equality matching.
4671 52. Exim was using ints instead of ino_t variables in some places where it was
4672 dealing with inode numbers.
4674 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4675 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4676 called TMPDIR, and if it finds it is different, it changes its value.
4678 54. The smtp_printf() function is now made available to local_scan() so
4679 additional output lines can be written before returning. There is also an
4680 smtp_fflush() function to enable the detection of a dropped connection.
4681 The variables smtp_input and smtp_batched_input are exported to
4684 55. Changed the default runtime configuration: the message "Unknown user"
4685 has been removed from the ACL, and instead placed on the localuser router,
4686 using the cannot_route_message feature. This means that any verification
4687 failures that generate their own messages won't get overridden. Similarly,
4688 the "Unrouteable address" message that was in the ACL for unverifiable
4689 relay addresses has also been removed.
4691 56. Added hosts_avoid_esmtp to the smtp transport.
4693 57. The exicyclog script was not checking for the esoteric option
4694 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4695 will work only if exicyclog is run under the appropriate euid.
4697 58. Following a discussion on the list, the rules by which Exim recognises line
4698 endings on incoming messages have been changed. The -dropcr and drop_cr
4699 options are now no-ops, retained only for backwards compatibility. The
4700 following line terminators are recognized: LF CRLF CR. However, special
4701 processing applies to CR:
4703 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4704 nor a local message in the state where . is a terminator.
4706 (ii) If a bare CR is encountered in a header line, an extra space is added
4707 after the line terminator so as not to end the header. The reasoning
4708 behind this is that bare CRs in header lines are most likely either
4709 to be mistakes, or people trying to play silly games.
4711 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4712 was being incorrectly given as 18 bytes larger than it should have been.
4713 This is a VOB (very old bug).
4715 60. This may never have affected anything current, but just in case it has:
4716 When the local host is found other than at the start of a list of hosts,
4717 the local host, those with the same MX, and any that follow, are discarded.
4718 When the list in question was part of a longer list of hosts, the following
4719 hosts (not currently being processed) were also being discarded. This no
4720 longer happens. I'm not sure if this situation could ever has previously
4723 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4726 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4727 header. This is something that is recommended in a new Internet Draft, and
4728 is something that is documented as being done by Sendmail. There are two
4729 possible values. For messages generated by the autoreply transport, Exim
4732 Auto-Submitted: auto-replied
4734 whereas for all other generated messages (e.g. bounces) it adds
4736 Auto-Submitted: auto-generated
4738 63. The "personal" condition in filters now includes a test for the
4739 Auto-Submitted: header. If it contains the string "auto-" the message it
4740 not considered personal.
4742 64. Added rcpt_include_affixes as a generic transport option.
4744 65. Added queue_only_override (default true).
4746 66. Added the syslog_duplication option.
4748 67. If what should have been the first header line of a message consisted of
4749 a space followed by a colon, Exim was mis-interpreting it as a header line.
4750 It isn't of course - it is syntactically invalid and should therefore be
4751 treated as the start of the message body. The misbehaviour could have
4752 caused a number of strange effects, including loss of data in subsequent
4753 header lines, and spool format errors.
4755 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4756 client host had authenticated. This control can now be exercised by an ACL
4757 for more flexibility.
4759 69. By default, callouts do not happen when testing with -bh. There is now a
4760 variant, -bhc, which does actually run the callout code, including
4761 consulting and updating the callout cache.
4763 70. Added support for saslauthd authentication, courtesy of Alexander
4766 71. If statvfs() failed on the spool or log directories while checking their
4767 size for availability, Exim confusingly gave the error "space shortage".
4768 Furthermore, in debugging mode it crashed with a floating point exception.
4769 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4770 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4771 serious problem, Exim now writes to the main and panic logs when this
4772 happens, with details of the failure. It then refuses to accept the
4773 incoming message, giving the message "spool directory problem" or "log
4774 directory problem" with a 421 code for SMTP messages.
4776 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4777 0, 1, and 2 exist, because some OS complain for execs without them (see
4778 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4779 descriptors. However, the code omitted to check that the open succeeded,
4780 causing mysterious errors if for some reason the permissions on /dev/null
4781 got screwed. Now Exim writes a message to the main and panic logs, and
4782 bombs out if it can't open /dev/null.
4784 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4785 interact so that it is all more flexible. It is supposed to remain
4786 backwards compatible. Also added extra_local_interfaces.
4788 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4789 to bomb out with an assertion failure - to the client this appears as a
4790 connection drop. This problem occurs in the part of the code that was taken
4791 from the Samba project. Fortunately, the assertion is in a very simple
4792 function, so I have fixed this by reproducing the function inline in the
4793 one place where it is called, and arranging for authentication to fail
4794 instead of killing the process with assert().
4796 75. The SPA client code was not working when the server requested OEM rather
4797 than Unicode encoding.
4799 76. Added code to make require_files with a specific uid setting more usable in
4800 the case where statting the file as root fails - usually a non-root-mounted
4801 NFS file system. When this happens and the failure is EACCES, Exim now
4802 forks a subprocess and does the per-uid checking as the relevant uid.
4804 77. Added process_log_path.
4806 78. If log_file_path was not explicitly set, a setting of check_log_space or
4807 check_log_inodes was ignored.
4809 79. If a space check for the spool or log partitions fails, the incident is now
4810 logged. Of course, in the latter case the data may get lost...
4812 80. Added the %p formatting code to string_format() so that it can be used to
4813 print addresses in debug_print(). Adjusted all the address printing in the
4814 debugging in store.c to use %p rather than %d.
4816 81. There was a concern that a line of code in smtp_in.c could overflow a
4817 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4818 initially expressed, the concern was not well-founded, because trailing
4819 spaces are removed early. However, if the trailing spaces were followed by
4820 a NULL, they did not get removed, so the overflow was possible. Two fixes
4823 (a) I re-wrote the offending code in a cleaner fashion.
4824 (b) If an incoming SMTP command contains a NULL character, it is rejected
4827 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4828 runs initgroups(), so that if the Exim user is in any additional groups,
4829 they will be used during message reception.
4835 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob