1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.518 2007/06/26 09:23:34 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.
78 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
79 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
82 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
85 PH/02 In an ACL statement such as
87 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
89 if a client was not listed at all, or was listed with a value other than
90 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
91 the condition was not true (as it should be), so access was not denied.
92 The bug was that the ! inversion was incorrectly passed on to the second
93 item. This has been fixed.
95 PH/03 Added additional dnslists conditions == and =& which are different from
96 = and & when the dns lookup returns more than one IP address.
98 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
99 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
101 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
102 FSYNC, which compiles an option called disable_fsync that allows for
103 bypassing fsync(). The documentation is heavily laced with warnings.
105 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
107 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
108 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
109 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
110 including adding "make clean"; (3) Added -fPIC when compiling the test
111 dynamically loaded module, to get rid of a warning.
113 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
114 message fails, move_frozen_messages = true and ignore_bounce_errors_after
115 = 0s. The bug is otherwise harmless.
117 PH/07 There was a bug in the dovecot authenticator such that the value of
118 $auth1 could be overwritten, and so not correctly preserved, after a
119 successful authentication. This usually meant that the value preserved by
120 the server_setid option was incorrect.
122 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
124 PH/09 Installed PCRE release 7.0.
126 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
127 run for batched SMTP input. It is now run at the start of every message
128 in the batch. While fixing this I discovered that the process information
129 (output by running exiwhat) was not always getting set for -bs and -bS
130 input. This is fixed, and it now also says "batched" for BSMTP.
132 PH/11 Added control=no_pipelining.
134 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
135 patch, slightly modified), and move the expansion of helo_data till after
136 the connection is made in the smtp transport (so it can use these
139 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
141 PH/14 Added log_selector = +pid.
143 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
145 PH/16 Add ${if forany and ${if forall.
147 PH/17 Added dsn_from option to vary the From: line in DSNs.
149 PH/18 Flush SMTP output before performing a callout, unless control =
150 no_callout_flush is set.
152 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
153 was true (the default) a successful delivery failed to delete the retry
154 item, thus causing premature timeout of the address. The bug is now
157 PH/20 Added hosts_avoid_pipelining to the smtp transport.
159 PH/21 Long custom messages for fakedefer and fakereject are now split up
160 into multiline reponses in the same way that messages for "deny" and
161 other ACL rejections are.
163 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
164 with slight modification.
166 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
167 draft, changing the syntax and factoring some duplicate code".
169 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
170 for deliveries of the second and subsequent messages over the same SMTP
173 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
174 ${reduce, with only minor "tidies".
176 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
178 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
179 expansion side effects.
181 PH/27 When a message times out after an over-quota error from an Exim-imposed
182 quota, the bounce message says "mailbox is full". This message was not
183 being given when it was a system quota that was exceeded. It now should
186 MH/03 Made $recipients available in local_scan(). local_scan() already has
187 better access to the recipient list through recipients_list[], but
188 $recipients can be useful in postmaster-provided expansion strings.
190 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
191 in the case of a MAIL command with additional options following the
192 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
193 were accidentally chopped off.
195 PH/29 SMTP synchronization checks are implemented when a command is read -
196 there is a check that no more input is waiting when there shouldn't be
197 any. However, for some commands, a delay in an ACL can mean that it is
198 some time before the response is written. In this time, more input might
199 arrive, invalidly. So now there are extra checks after an ACL has run for
200 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
201 pipelining has not been advertised.
203 PH/30 MH's patch to allow iscntrl() characters to be list separators.
205 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
206 returned in the SMTP response when smtp_return_error_details was false.
209 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
210 instead of the C I/O that was originally supplied, because problems were
213 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
214 Exim which did not show up earlier: it was assuming that a call to
215 SSL_CTX_set_info_callback() might give an error value. In fact, there is
216 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
217 was a macro that became an assignment, so it seemed to work. This has
218 changed to a proper function call with a void return, hence the compile
219 error. Exim's code has been fixed.
221 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
224 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
226 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
228 PH/37 If a message is not accepted after it has had an id assigned (e.g.
229 because it turns out to be too big or there is a timeout) there is no
230 "Completed" line in the log. When some messages of this type were
231 selected by exigrep, they were listed as "not completed". Others were
232 picked up by some special patterns. I have improved the selection
233 criteria to be more general.
235 PH/38 The host_find_failed option in the manualroute router can now be set
236 to "ignore", to completely ignore a host whose IP address cannot be
237 found. If all hosts are ignored, the behaviour is controlled by the new
238 host_all_ignored option.
240 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
241 homing or because of multiple MX records with /mx) generated more than
242 one IP address, and the following item turned out to be the local host,
243 all the secondary addresses of the first item were incorrectly removed
244 from the list, along with the local host and any following hosts (which
245 is what is supposed to happen).
247 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
248 whoever called Exim into the -H file. In the case of the daemon it was
249 behaving confusingly. When first started, it used values for whoever
250 started the daemon, but after a SIGHUP it used the Exim user (because it
251 calls itself on a restart). I have changed the code so that it now always
254 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
255 message are rejected with the same error (e.g. no authentication or bad
256 sender address), and a DATA command is nevertheless sent (as can happen
257 with PIPELINING or a stupid MUA), the error message that was given to the
258 RCPT commands is included in the rejection of the DATA command. This is
259 intended to be helpful for MUAs that show only the final error to their
262 PH/42 Another patch from the Sieve maintainer.
264 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
265 Eximstats - Fixed some broken HTML links and added missing column headers
267 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
268 columns for Rejects, Temp Rejects, Ham, and Spam rows.
270 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
272 PH/43 Yet another patch from the Sieve maintainer.
274 PH/44 I found a way to check for a TCP/IP connection going away before sending
275 the response to the final '.' that terminates a message, but only in the
276 case where the client has not sent further data following the '.'
277 (unfortunately, this is allowed). However, in many cases there won't be
278 any further data because there won't be any more messages to send. A call
279 to select() can be used: if it shows that the input is "ready", there is
280 either input waiting, or the socket has been closed. An attempt to read
281 the next input character can distinguish the two cases. Previously, Exim
282 would have sent an OK response which the client would never have see.
283 This could lead to message repetition. This fix should cure that, at
284 least in a lot of common cases.
286 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
287 advertised in response to EHLO.
293 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
294 fixed by 4.65/MH/01 (is this a record?) are fixed:
296 (i) An empty string was always treated as zero by the numeric comparison
297 operators. This behaviour has been restored.
299 (ii) It is documented that the numeric comparison operators always treat
300 their arguments as decimal numbers. This was broken in that numbers
301 starting with 0 were being interpreted as octal.
303 While fixing these problems I realized that there was another issue that
304 hadn't been noticed. Values of message_size_limit (both the global option
305 and the transport option) were treated as octal if they started with 0.
306 The documentation was vague. These values are now always treated as
307 decimal, and I will make that clear in the documentation.
313 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
314 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
317 MH/01 Don't check that the operands of numeric comparison operators are
318 integers when their expansion is in "skipping" mode (fixes bug
319 introduced by 4.64-PH/07).
321 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
322 child addresses, Exim now panics and dies. Previously, because the count
323 is held in a short int, deliveries were likely to be lost. As such a
324 large number of recipients for a single message is ridiculous
325 (performance will be very, very poor), I have chosen to impose a limit
326 rather than extend the field.
332 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
333 leftover -K file (the existence of which was triggered by #402).
334 While we were at it, introduced process PID as part of the -K
335 filename. This should rule out race conditions when creating
338 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
339 processing considerably. Previous code took too long for large mails,
340 triggering a timeout which in turn triggers #401.
342 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
343 in the DK code in transports.c. sendfile() is not really portable,
344 hence the _LINUX specificness.
346 TF/01 In the add_headers option to the mail command in an Exim filter,
347 there was a bug that Exim would claim a syntax error in any
348 header after the first one which had an odd number of characters
351 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
352 callout verification, Exim cached a "reject" for the entire domain. This
353 is correct for most verifications, but it is not correct for a recipient
354 verification with use_sender or use_postmaster set, because in that case
355 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
356 case of MAIL FROM:<> rejection from other early rejections (e.g.
357 rejection of HELO). When verifying a recipient using a non-null MAIL
358 address, the cache is ignored if it shows MAIL FROM:<> rejection.
359 Whatever the result of the callout, the value of the domain cache is
360 left unchanged (for any other kind of callout, getting as far as trying
361 RCPT means that the domain itself is ok).
363 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
366 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
367 manifest itself as EPIPE rather than ECONNECT. When tidying away a
368 session, the daemon ignores ECONNECT errors and logs others; it now
369 ignores EPIPE as well.
371 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
372 (quoted-printable decoding).
374 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
375 later the small subsequent patch to fix an introduced bug.
377 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
379 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
381 PH/08 An error is now given if message_size_limit is specified negative.
383 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
384 to be given (somewhat) arbitrary names.
386 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
389 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
390 miscellaneous code fixes
392 PH/10 Added the log_reject_target ACL modifier to specify where to log
395 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
396 hostname. This is wrong, because it relates to the incoming message (and
397 probably the interface on which it is arriving) and not to the outgoing
398 callout (which could be using a different interface). This has been
399 changed to use the value of the helo_data option from the smtp transport
400 instead - this is what is used when a message is actually being sent. If
401 there is no remote transport (possible with a router that sets up host
402 addresses), $smtp_active_hostname is used.
404 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
405 tweaks were necessary in order to get it to work (see also 21 below):
406 (a) The code assumed that strncpy() returns a negative number on buffer
407 overflow, which isn't the case. Replaced with Exim's string_format()
409 (b) There were several signed/unsigned issues. I just did the minimum
410 hacking in of casts. There is scope for a larger refactoring.
411 (c) The code used strcasecmp() which is not a standard C function.
412 Replaced with Exim's strcmpic() function.
413 (d) The code set only $1; it now sets $auth1 as well.
414 (e) A simple test gave the error "authentication client didn't specify
415 service in request". It would seem that Dovecot has changed its
416 interface. Fortunately there's a specification; I followed it and
417 changed what the client sends and it appears to be working now.
419 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
422 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
423 address A is aliased to B and C, where B exists and C does not. Without
424 -v the output is "A verified" because verification stops after a
425 successful redirection if more than one address is generated. However,
426 with -v the child addresses are also verified. Exim was outputting "A
427 failed to verify" and then showing the successful verification for C,
428 with its parentage. It now outputs "B failed to verify", showing B's
429 parentage before showing the successful verification of C.
431 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
432 look up a TXT record in a specific list after matching in a combined
435 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
436 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
437 they consult the DNS. I had assumed they would set it the way they
438 wanted; and indeed my experiments on Linux seem to show that in some
439 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
440 To be on the safe side, however, I have now made the interface to
441 host_find_byname() similar to host_find_bydns(), with an argument
442 containing the DNS resolver options. The host_find_byname() function now
443 sets these options at its start, just as host_find_bydns() does. The smtp
444 transport options dns_qualify_single and dns_search_parents are passed to
445 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
446 of host_find_byname() use the default settings of RES_DEFNAMES
447 (qualify_single) but not RES_DNSRCH (search_parents).
449 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
450 spool_read_header() do less string testing, by means of a preliminary
451 switch on the second character of optional "-foo" lines. (This is
452 overdue, caused by the large number of possibilities that now exist.
453 Originally there were few.) While I was there, I also converted the
454 str(n)cmp tests so they don't re-test the leading "-" and the first
455 character, in the hope this might squeeze out yet more improvement.
457 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
458 flag allowing group syntax was set by the header_syntax check but not
459 turned off, possible causing trouble later; (2) The flag was not being
460 set at all for the header_verify test, causing "group"-style headers to
461 be rejected. I have now set it in this case, and also caused header_
462 verify to ignore an empty address taken from a group. While doing this, I
463 came across some other cases where the code for allowing group syntax
464 while scanning a header line wasn't quite right (mostly, not resetting
465 the flag correctly in the right place). These bugs could have caused
466 trouble for malformed header lines. I hope it is now all correct.
468 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
469 with the "reply" argument non-NULL. The code, however (which originally
470 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
471 but it didn't always do it. This confused somebody who was copying the
472 code for some other use. I have removed all the tests.
474 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
475 feature that was used to support insecure browsers during the U.S. crypto
476 embargo. It requires special client support, and Exim is probably the
477 only MTA that supported it -- and would never use it because real RSA is
478 always available. This code has been removed, because it had the bad
479 effect of slowing Exim down by computing (never used) parameters for the
480 RSA_EXPORT functionality.
482 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
483 authenticator to fail if there's a tab character in the incoming data
484 (there should never be unless someone is messing about, as it's supposed
485 to be base64-encoded). Also added, on Timo's advice, the "secured" option
486 if the connection is using TLS or if the remote IP is the same as the
487 local IP, and the "valid-client-cert option" if a client certificate has
490 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
491 authenticators. This can be used for authorization after authentication
492 succeeds. (In the case of plaintext, it servers for both authentication
495 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
496 if any retry times were supplied.
498 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
499 connection, where there is no client IP address to check. In this
500 situation, the verify now always succeeds.
502 PH/25 Applied John Jetmore's -Mset patch.
504 PH/26 Added -bem to be like -Mset, but loading a message from a file.
506 PH/27 In a string expansion for a processed (not raw) header when multiple
507 headers of the same name were present, leading whitespace was being
508 removed from all of them, but trailing whitespace was being removed only
509 from the last one. Now trailing whitespace is removed from each header
510 before concatenation. Completely empty headers in a concatenation (as
513 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
514 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
516 PH/29 [Removed. This was a change that I later backed out, and forgot to
517 correct the ChangeLog entry (that I had efficiently created) before
518 committing the later change.]
520 PH/30 Exim was sometimes attempting to deliver messages that had suffered
521 address errors (4xx response to RCPT) over the same connection as other
522 messages routed to the same hosts. Such deliveries are always "forced",
523 so retry times are not inspected. This resulted in far too many retries
524 for the affected addresses. The effect occurred only when there were more
525 hosts than the hosts_max_try setting in the smtp transport when it had
526 the 4xx errors. Those hosts that it had tried were not added to the list
527 of hosts for which the message was waiting, so if all were tried, there
528 was no problem. Two fixes have been applied:
530 (i) If there are any address or message errors in an SMTP delivery, none
531 of the hosts (tried or untried) are now added to the list of hosts
532 for which the message is waiting, so the message should not be a
533 candidate for sending over the same connection that was used for a
534 successful delivery of some other message. This seems entirely
535 reasonable: after all the message is NOT "waiting for some host".
536 This is so "obvious" that I'm not sure why it wasn't done
537 previously. Hope I haven't missed anything, but it can't do any
538 harm, as the worst effect is to miss an optimization.
540 (ii) If, despite (i), such a delivery is accidentally attempted, the
541 routing retry time is respected, so at least it doesn't keep
542 hammering the server.
544 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
545 in ${readsocket because some servers need this prod.
547 PH/32 Added some extra debug output when updating a wait-xxx database.
549 PH/33 The hint "could be header name not terminated by colon", which has been
550 given for certain expansion errors for a long time, was not being given
551 for the ${if def:h_colon_omitted{... case.
553 PH/34 The spec says: "With one important exception, whenever a domain list is
554 being scanned, $domain contains the subject domain." There was at least
555 one case where this was not true.
557 PH/35 The error "getsockname() failed: connection reset by peer" was being
558 written to the panic log as well as the main log, but it isn't really
559 panic-worthy as it just means the connection died rather early on. I have
560 removed the panic log writing for the ECONNRESET error when getsockname()
563 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
564 runs only) independently of the message's sender address. This meant
565 that, if the 4xx error was in fact related to the sender, a different
566 message to the same recipient with a different sender could confuse
567 things. In particualar, this can happen when sending to a greylisting
568 server, but other circumstances could also provoke similar problems.
569 I have changed the default so that the retry time for these errors is now
570 based a combination of the sender and recipient addresses. This change
571 can be overridden by setting address_retry_include_sender=false in the
574 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
575 remote server are returned as part of bounce messages. This was not
576 happening for LMTP over a pipe (the lmtp transport), but now it is the
577 same for both kinds of LMTP.
579 PH/38 Despite being documented as not happening, Exim was rewriting addresses
580 in header lines that were in fact CNAMEs. This is no longer the case.
582 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
583 and queue runs started by the daemon processed all messages. This has
584 been fixed so that -R and -S can now usefully be given with -q<time>.
586 PH/40 Import PCRE release 6.7 (fixes some bugs).
588 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
590 PH/42 Give an error if -q is specified more than once.
592 PH/43 Renamed the variables $interface_address and $interface_port as
593 $received_ip_address and $received_port, to make it clear that these
594 values apply to message reception, and not to the outgoing interface when
595 a message is delivered. (The old names remain recognized, of course.)
597 PH/44 There was no timeout on the connect() call when using a Unix domain
598 socket in the ${readsocket expansion. There now is.
600 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
601 be meaningful with "accept".
603 SC/01 Eximstats V1.43
604 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
606 SC/02 Eximstats V1.44
607 Use a glob alias rather than an array ref in the generated
608 parser. This improves both readability and performance.
610 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
611 Collect SpamAssassin and rejection statistics.
612 Don't display local sender or destination tables unless
613 there is data to show.
614 Added average volumes into the top table text output.
616 SC/04 Eximstats V1.46
617 Collect data on the number of addresses (recipients)
618 as well as the number of messages.
620 SC/05 Eximstats V1.47
621 Added 'Message too big' to the list of mail rejection
622 reasons (thanks to Marco Gaiarin).
624 SC/06 Eximstats V1.48
625 Mainlog lines which have GMT offsets and are too short to
626 have a flag are now skipped.
628 SC/07 Eximstats V1.49 (Alain Williams)
629 Added the -emptyok flag.
631 SC/08 Eximstats V1.50
632 Fixes for obtaining the IP address from reject messages.
634 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
635 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
636 whitesspace changes from 4.64-PH/27
638 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
641 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
642 are found, allow negative numbers in numeric criteria)
644 JJ/06 exipick.20061117.2, added new $message_body_missing variable
646 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
647 to match changes made in 4.64-PH/43
649 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
651 PH/47 Put in an explicit test for a DNS lookup of an address record where the
652 "domain" is actually an IP address, and force a failure. This locks out
653 those revolvers/nameservers that support "A-for-A" lookups, in
654 contravention of the specifications.
656 PH/48 When a host name was looked up from an IP address, and the subsequent
657 forward lookup of the name timed out, the host name was left in
658 $sender_host_name, contrary to the specification.
660 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
661 restricted to single-key lookups, Exim was not diagnosing an error if
662 * or *@ was used with a query-style lookup.
664 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
666 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
667 long ago, but noone interested enough thought of it. Let's just say that
668 the "1.1" means that there are some new functions that weren't there at
669 some point in the past.
671 PH/51 Error processing for expansion failure of helo_data from an smtp
672 transport during callout processing was broken.
674 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
675 tested/used via the -bh/-bhc/-bs options.
677 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
678 bug, fixed in subsequent PCRE releases).
680 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
681 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
683 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
689 SC/01 Use a glob alias rather than an array ref in eximstats generated
690 parser. This improves both readability and performance.
692 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
693 Don't display local sender or destination tables in eximstats unless
694 there is data to show.
695 Added average volumes into the eximstats top table text output.
697 SC/03 Collect data on the number of addresses (recipients) as well
698 as the number of messages in eximstats.
700 TF/01 Correct an error in the documentation for the redirect router. Exim
701 does (usually) call initgroups() when daemonizing.
703 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
704 with consistent privilege compared to when running as a daemon.
706 TF/03 Note in the spec that $authenticated_id is not set for local
707 submissions from trusted users.
709 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
710 Thanks to Dean Brooks <dean@iglou.com> for the patch.
712 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
713 by adding some example configuration directives to the default
714 configuration file. A little bit of work is required to uncomment the
715 directives and define how usernames and passwords are checked, but
716 there is now a framework to start from.
718 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
719 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
720 without this. I don't know how relevant this is to other LDAP libraries.
722 PH/02 Add the verb name to the "unknown ACL verb" error.
724 PH/03 Magnus Holmgren's patch for filter_prepend_home.
726 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
728 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
729 directory not expanded when it should be if an expanded home directory
730 was set for the address (which is overridden by the transport).
732 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
735 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
736 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
737 because it is too late at that time, and has no effect.
739 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
740 security issue with \' (bugzilla #107). I could not use the
741 PQescapeStringConn() function, because it needs a PGconn value as one of
744 PH/08 When testing addresses using -bt, indicate those final addresses that
745 are duplicates that would not cause an additional delivery. At least one
746 person was confused, thinking that -bt output corresponded to deliveries.
747 (Suppressing duplicates isn't a good idea as you lose the information
748 about possibly different redirections that led to the duplicates.)
750 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
751 systems where poll() doesn't work, in particular OS X.
753 PH/10 Added more information to debugging output for retry time not reached.
755 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
756 operations in malware.c.
758 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
761 PH/13 If write_rejectlog was set false when logging was sent to syslog with
762 syslog_duplication set false, log lines that would normally be written
763 both the the main log and to the reject log were not written to syslog at
766 PH/14 In the default configuration, change the use of "message" in ACL warn
767 statements to "add_header".
769 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
770 not followed by a command (e.g. "seen endif").
772 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
773 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
776 PH/17 Added extra conditions to the default value of delay_warning_condition
780 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
781 { match{$h_precedence:}{(?i)bulk|list|junk} } \
782 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
785 The Auto-Submitted: and various List- headers are standardised, whereas I
786 don't think Precedence: ever was.
788 PH/18 Refactored debugging code in route_finduser() to show more information,
789 in particular, the error code if getpwnam() issues one.
791 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
792 This is apparently needed in addition to the PH/07 change above to avoid
793 any possible encoding problems.
795 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
796 but not after initializing Perl.
798 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
799 output them only if debugging. By default they are written stderr,
800 apparently, which is not desirable.
802 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
805 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
808 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
810 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
811 authenticated or an ident call has been made. Suppress the default
812 values for $authenticated_id and $authenticated_sender (but permit -oMai
813 and -oMas) when testing with -bh.
815 PH/24 Re-jigged the order of the tests in the default configuration so that the
816 tests for valid domains and recipients precede the DNS black list and CSA
817 tests, on the grounds that those ones are more expensive.
819 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
820 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
821 This bug exists in every version of Exim that I still have, right back to
824 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
825 However, an attempt to turn on case-sensitivity in a regex key by
826 including (?-i) didn't work because the subject string was already
827 lowercased, and the effects were non-intuitive. It turns out that a
828 one-line patch can be used to allow (?-i) to work as expected.
834 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
835 other effects) broke the use of negated acl sub-conditions.
837 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
840 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
841 "Deny" causes Exim to reject the incoming connection with a 554 error.
842 Unfortunately, if there is a major crisis, such as a disk failure,
843 tcp-wrappers gives "deny", whereas what one would like would be some
844 kind of temporary error. A kludge has been added to help with this.
845 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
846 554 error is used if errno is still zero or contains ENOENT (which occurs
847 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
850 PH/03 Add -lutil to the default FreeBSD LIBS setting.
852 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
853 errors. Otherwise a message that provokes a temporary error (when other
854 messages do not) can cause a whole host to time out.
856 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
857 addresses were routed directly to files or pipes from a redirect router.
858 File deliveries just didn't batch; pipe deliveries might have suffered
861 PH/06 A failure to get a lock for a hints database would erroneously always say
862 "Failed to get write lock", even when it was really a read lock.
864 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
865 of 0600. This has been changed to use the value of the lockfile_mode
866 option (which defaults to 0600).
868 PH/08 Applied small patch from the Sieve maintainer.
870 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
871 folder from quota calculations, a direct delivery into this folder messed
872 up the contents of the maildirsize file. This was because the regex was
873 used only to exclude .Trash (or whatever) when the size of the mailbox
874 was calculated. There was no check that a delivery was happening into an
875 excluded directory. This bug has been fixed by ignoring all quota
876 processing for deliveries into excluded directories.
878 PH/10 Added the maildirfolder_create_regex option to appendfile.
884 PH/01 The code for finding all the local interface addresses on a FreeBSD
885 system running IPv6 was broken. This may well have applied to all BSD
886 systems, as well as to others that have similar system calls. The broken
887 code found IPv4 interfaces correctly, but gave incorrect values for the
888 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
889 that it would not match correctly against @[] and not recognize the IPv6
892 PH/02 The ipliteral router was not recognizing addresses of the form user@
893 [ipv6:....] because it didn't know about the "ipv6:" prefix.
895 PH/03 Added disable_ipv6.
897 PH/04 Changed $reply_address to use the raw form of the headers instead of the
898 decoded form, because it is most often used to construct To: headers
899 lines in autoreplies, and the decoded form may well be syntactically
900 invalid. However, $reply_address has leading white space removed, and all
901 newlines turned into spaces so that the autoreply transport does not
904 PH/05 If group was specified without a user on a router, and no group or user
905 was specified on a transport, the group from the router was ignored.
907 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
908 for visible compile-time settings that can be used to change these
909 numbers, for those that want even more. Backwards compatibility with old
910 spool files has been maintained. However, going back to a previous Exim
911 release will lost any variables that are in spool files.
913 PH/07 Two small changes when running in the test harness: increase delay when
914 passing a TCP/IP connection to a new process, in case the original
915 process has to generate a bounce, and remove special handling of
916 127.0.0.2 (sic), which is no longer necessary.
918 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
919 be the same on different OS.
921 PH/09 Moved a debug statement in filter processing to avoid a race problem when
924 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
925 whether --show-vars was specified or not
927 JJ/02 exipick: Added support for new ACL variable spool format introduced
930 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
931 syntactically invalid From: or Reply-to: line, and a filter used this to
932 generate an autoreply, and therefore failed to obtain an address for the
933 autoreply, Exim could try to deliver to a non-existent relative file
934 name, causing unrelated and misleading errors. What now happens is that
935 it logs this as a hard delivery error, but does not attempt to create a
938 PH/11 The exinext utility has a -C option for testing purposes, but although
939 the given file was scanned by exinext itself; it wasn't being passed on
940 when Exim was called.
942 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
943 an end-of-file indication when reading a command response.
945 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
946 compiled. In many other places in Exim, IPv6 addresses are always
947 recognized, so I have changed this. It also means that IPv4 domain
948 literals of the form [IPV4:n.n.n.n] are now always recognized.
950 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
951 used if the router is not running as root, for example, when verifying at
952 ACL time, or when using -bh. The debugging output from this situation was
953 non-existent - all you got was a failure to exec. I have made two
956 (a) Failures to set uid/gid, the current directory, or a process leader
957 in a subprocess such as that created by queryprogram now generate
958 suitable debugging ouput when -d is set.
960 (b) The queryprogram router detects when it is not running as root,
961 outputs suitable debugging information if -d is set, and then runs
962 the subprocess without attempting to change uid/gid.
964 PH/15 Minor change to Makefile for building test_host (undocumented testing
967 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
968 additional section of a DNS packet that returns MX or SRV records.
969 Instead, it always explicitly searches for A/AAAA records. This avoids
970 major problems that occur when a DNS server includes only records of one
971 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
972 fixed another bug: if SRV records were looked up and the corresponding
973 address records were *not* found in the additional section, the port
974 values from the SRV records were lost.
976 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
977 using the correct key (the original address) when searching the retry
978 rules in order to find which one to use for generating the retry hint.
980 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
981 adding the default one. Similarly, if it contains a Reply-To: header, the
982 errors_reply_to option, if set, is not used.
984 PH/19 When calculating a retry time, Exim used to measure the "time since
985 failure" by looking at the "first failed" field in the retry record. Now
986 it does not use this if it is later than than the arrival time of the
987 message. Instead it uses the arrival time. This makes for better
988 behaviour in cases where some deliveries succeed, thus re-setting the
989 "first failed" field. An example is a quota failure for a huge message
990 when small messages continue to be delivered. Without this change, the
991 "time since failure" will always be short, possible causing more frequent
992 delivery attempts for the huge message than are intended.
993 [Note: This change was subsequently modified - see PH/04 for 4.62.]
995 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
996 $1, $2, $3) because the numerical variables can be reset during some
997 expansion items (e.g. "match"), thereby losing the authentication data.
999 PH/21 Make -bV show the size of off_t variables so that the test suite can
1000 decide whether to run tests for quotas > 2G.
1002 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1003 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1004 filecount value is greater than 2G or if a quota value is greater than 2G
1005 on a system where the size of off_t is not greater than 4, a panic error
1008 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1009 never match. The debug and -bh output now contains an explicit error
1010 message indicating a malformed IPv4 address or mask.
1012 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1013 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1014 PH/23 above applies.
1016 PH/25 Do not write to syslog when running in the test harness. The only
1017 occasion when this arises is a failure to open the main or panic logs
1018 (for which there is an explicit test).
1020 PH/26 Added the /no_tell option to "control=freeze".
1022 PH/27 If a host name lookup failed very early in a connection, for example, if
1023 the IP address matched host_lookup and the reverse lookup yielded a name
1024 that did not have a forward lookup, an error message of the form "no IP
1025 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1026 could be logged. Now it outputs the IP address instead of "NULL".
1028 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1029 allows the sender and the authenticated sender to be set when
1030 submitting a message from within Exim. Since child_open_exim() is
1031 documented for local_scan(), the new function should be too.
1033 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1034 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1035 results in an empty string is now treated as unset.
1037 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1039 PH/31 Added sender_verify_fail logging option.
1041 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1042 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1043 tidied the source and removed it altogether.
1045 PH/33 When a queue run was abandoned because the load average was too high, a
1046 log line was always written; now it is written only if the queue_run log
1047 selector is set. In addition, the log line for abandonment now contains
1048 information about the queue run such as the pid. This is always present
1049 in "start" and "stop" lines but was omitted from the "abandon" line.
1051 PH/34 Omit spaces between a header name and the colon in the error message that
1052 is given when verify = headers_syntax fails (if there are lots of them,
1053 the message gets confusing).
1055 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1056 names, as there are now some PTR records that contain slashes. This check
1057 is only to protect against broken name servers that fall over on strange
1058 characters, so the fact that it applies to all lookups doesn't matter.
1060 PH/36 Now that the new test suite is complete, we can remove some of the
1061 special code in Exim that was needed for the old test suite. For example,
1062 sorting DNS records because real resolvers return them in an arbitrary
1063 order. The new test suite's fake resolver always returns records in the
1066 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1067 bounces) except when queue_only is set, to avoid logging races between
1068 the different processes.
1070 PH/38 Panic-die if .include specifies a non-absolute path.
1072 PH/39 A tweak to the "H" retry rule from its user.
1074 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1075 a label. They prevented compilation on older perls.
1077 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1078 a warning to be raised on newish perls.
1080 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1081 on queue. Changes to match documented behaviour of showing count of
1082 messages matching specified criteria.
1084 PH/40 Changed the default ident timeout from 30s to 5s.
1086 PH/41 Added support for the use of login_cap features, on those BSD systems
1087 that have them, for controlling the resources used by pipe deliveries.
1089 PH/42 The content-scanning code uses fopen() to create files in which to put
1090 message data. Previously it was not paying any attention to the mode of
1091 the files. Exim runs with umask(0) because the rest of the code creates
1092 files with open(), and sets the required mode explicitly. Thus, these
1093 files were ending up world-writeable. This was not a big issue, because,
1094 being within the spool directory, they were not world-accessible. I have
1095 created a function called modefopen, which takes an additional mode
1096 argument. It sets umask(777), creates the file, chmods it to the required
1097 mode, then resets the umask. All the relevant calls to fopen() in the
1098 content scanning code have been changed to use this function.
1100 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1101 to 24 hours. This avoids potential overflow problems when processing G
1102 and H retry rules. I suspect nobody ever tinkers with this value.
1104 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1106 PH/45 When the plaintext authenticator is running as a client, the server's
1107 challenges are checked to ensure they are valid base64 strings. By
1108 default, the authentication attempt is cancelled if an invalid string is
1109 received. Setting client_ignore_invalid_base64 true ignores these errors.
1110 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1111 they are received. Thus, the responses can be made to depend on the
1112 challenges. If an invalid string is ignored, an empty string is placed in
1115 PH/46 Messages that are created by the autoreply transport now contains a
1116 References: header, in accordance with RFCs 2822 and 3834.
1118 PH/47 Added authenticated_sender_force to the smtp transport.
1120 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1122 PH/49 Installed latest patch from the Sieve maintainer.
1124 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1125 was also set, the appendfile transport was unnecessarily scanning a
1126 directory of message files (e.g. for maildir delivery) to find the count
1127 of files (along with the size), even though it did not need this
1128 information. It now does the scan only if it needs to find either the
1129 size of the count of files.
1131 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1133 PH/52 Two bugs concerned with error handling when the smtp transport is
1136 (i) Exim was not creating retry information for temporary errors given
1137 for individual recipients after the DATA command when the smtp transport
1138 was used in LMTP mode. This meant that they could be retried too
1139 frequently, and not timed out correctly.
1141 (ii) Exim was setting the flag that allows error details to be returned
1142 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1143 recipients that were returned after the DATA command.
1145 PH/53 This is related to PH/52, but is more general: for any failing address,
1146 when detailed error information was permitted to be returned to the
1147 sender, but the error was temporary, then after the final timeout, only
1148 "retry timeout exceeded" was returned. Now it returns the full error as
1149 well as "retry timeout exceeded".
1151 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1152 do this, and (what is worse) MTAs that accept it.
1154 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1155 will now be deprecated.
1157 PH/56 New os.c-cygwin from the Cygwin maintainer.
1159 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1160 formats (previously only available in exim formats via -bpr, -bpru,
1161 and -bpra. Now also available in native and exiqgrep formats)
1163 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1164 with very large, slow to parse queues
1166 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1168 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1170 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1171 responses to authentication challenges, though it was showing the
1172 challenges; (ii) I've removed the CR characters from the debug output for
1175 PH/58 Allow for the insertion of a newline as well as a space when a string
1176 is turned into more than one encoded-word during RFC 2047 encoding. The
1177 Sieve code now uses this.
1179 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1180 data_4xx, lost_connection, tls_required.
1182 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1183 message was being sent as an SMTP response.
1185 PH/61 Add -l and -k options to exicyclog.
1187 PH/62 When verifying, if an address was redirected to one new address, so that
1188 verification continued, and the new address failed or deferred after
1189 having set something in $address_data, the value of $address_data was not
1190 passed back to the ACL. This was different to the case when no
1191 redirection occurred. The value is now passed back in both cases.
1193 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1194 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1195 use login_cap.h, so on its own it isn't the distinguishing feature. The
1196 new name refers directly to the setclassresources() function.
1198 PH/65 Added configuration files for NetBSD3.
1200 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1202 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1203 is preferred over IPv4.
1205 PH/68 The bounce_return_message and bounce_return_body options were not being
1206 honoured for bounces generated during the reception of non-SMTP messages.
1207 In particular, this applied to messages rejected by the ACL. This bug has
1208 been fixed. However, if bounce_return_message is true and bounce_return_
1209 body is false, the headers that are returned for a non-SMTP message
1210 include only those that have been read before the error was detected.
1211 (In the case of an ACL rejection, they have all been read.)
1213 PH/69 The HTML version of the specification is now built in a directory called
1214 spec_html instead of spec.html, because the latter looks like a path with
1215 a MIME-type, and this confuses some software.
1217 PH/70 Catch two compiler warnings in sieve.c.
1219 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1220 function verify_get_ident() calls ip_connect() to connect a socket, but
1221 if the "connect()" function timed out, ip_connect() used to close the
1222 socket. However, verify_get_ident() also closes the socket later, and in
1223 between Exim writes to the log, which may get opened at this point. When
1224 the socket was closed in ip_connect(), the log could get the same file
1225 descriptor number as the socket. This naturally causes chaos. The fix is
1226 not to close the socket in ip_connect(); the socket should be closed by
1227 the function that creates it. There was only one place in the code where
1228 this was missing, in the iplookup router, which I don't think anybody now
1229 uses, but I've fixed it anyway.
1231 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1232 well as to direct DNS lookups. Otherwise the handling of names in host
1233 lists is inconsistent and therefore confusing.
1239 PH/01 Two changes to the default runtime configuration:
1241 (1) Move the checks for relay_from_hosts and authenticated clients from
1242 after to before the (commented out) DNS black list checks.
1244 (2) Add control=submission to the relay_from_hosts and authenticated
1245 clients checks, on the grounds that messages accepted by these
1246 statements are most likely to be submissions.
1248 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1250 (1) Generate an error if the third argument for the ${prvs expansion is
1253 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1256 (3) Reset the variables that are obtained from the first argument of
1257 ${prvscheck and used in the second argument before leaving the code,
1258 because their memory is reclaimed, so using them afterwards may do
1261 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1262 one (it's much easier than Tom thought :-).
1264 (5) Because of (4), we can now allow for the use of $prvscheck_result
1265 inside the third argument.
1267 PH/03 For some reason, the default setting of PATH when running a command from
1268 a pipe transport was just "/usr/bin". I have changed it to
1271 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1272 anything to be listed in the output from -bV.
1274 PH/05 When a filter generated an autoreply, the entire To: header line was
1275 quoted in the delivery log line, like this:
1277 => >A.N.Other <ano@some.domain> <original@ddress> ...
1279 This has been changed so that it extracts the operative address. There
1280 may be more than one such address. If so, they are comma-separated, like
1283 => >ano@some.domain,ona@other.domain <original@ddress> ...
1285 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1286 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1287 not being looked up in the rDNS to get a host name, Exim was showing the
1288 IP address twice in Received: lines, even though the IP addresses were
1289 identical. For example:
1291 Received: from [1.2.3.4] (helo=[1.2.3.4])
1293 However, if the real host name was known, it was omitting the HELO data
1294 if it matched the actual IP address. This has been tidied up so that it
1295 doesn't show the same IP address twice.
1297 PH/07 When both +timestamp and +memory debugging was on, the value given by
1298 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1299 called by the debug printing, thereby overwriting the timestamp buffer.
1300 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1302 PH/08 When the original message was included in an autoreply transport, it
1303 always said "this is a copy of the message, including all the headers",
1304 even if body_only or headers_only was set. It now gives an appropriate
1307 PH/09 Applied a patch from the Sieve maintainer which:
1309 o fixes some comments
1310 o adds the (disabled) notify extension core
1311 o adds some debug output for the result of if/elsif tests
1312 o points to the current vacation draft in the documentation
1313 and documents the missing references header update
1317 o fixes a bug in processing the envelope test (when testing
1318 multiple envelope elements, the last element determinted the
1321 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1322 Electronic Mail") by including:
1324 Auto-submitted: auto-generated
1326 in the messages that it generates (bounce messages and others, such as
1327 warnings). In the case of bounce messages for non-SMTP mesages, there was
1328 also a typo: it was using "Auto_submitted" (underscore instead of
1329 hyphen). Since every message generated by Exim is necessarily in response
1330 to another message, thes have all been changed to:
1332 Auto-Submitted: auto-replied
1334 in accordance with these statements in the RFC:
1336 The auto-replied keyword:
1338 - SHOULD be used on messages sent in direct response to another
1339 message by an automatic process,
1341 - MUST NOT be used on manually-generated messages,
1343 - MAY be used on Delivery Status Notifications (DSNs) and Message
1344 Disposition Notifications (MDNs),
1346 - MUST NOT be used on messages generated by automatic or periodic
1347 processes, except for messages which are automatic responses to
1350 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1351 to the default Received: header definition.
1353 PH/12 Added log selector acl_warn_skipped (default on).
1355 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1356 variables because (a) they are in the wrong storage pool and (b) even if
1357 they were copied, it wouldn't work properly because of the caching.
1359 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1360 checking when decoding. Apparently there are clients that generate
1361 overlong encoded strings. Why am I not surprised?
1363 PH/15 If the first argument of "${if match_address" was not empty, but did not
1364 contain an "@" character, Exim crashed. Now it writes a panic log message
1365 and treats the condition as false.
1367 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1369 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1370 extension "envelope-auth". The code is finished and in agreement with
1371 other implementations, but there is no documentation so far and in fact,
1372 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1373 not changing the active code.
1375 Print executed "if" and "elsif" statements when debugging is used. This
1376 helps a great deal to understand what a filter does.
1378 Document more things not specified clearly in RFC3028. I had all this
1379 sorted out, when out of a sudden new issues came to my mind. Oops."
1381 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1384 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1385 canonical form (as documented). However, after a host name lookup from
1386 the IP address, check_host() was doing a simple string comparison with
1387 addresses acquired from the DNS when checking that the found name did
1388 have the original IP as one of its addresses. Since any found IPv6
1389 addresses are likely to be in abbreviated form, the comparison could
1390 fail. Luckily, there already exists a function for doing the comparison
1391 by converting both addresses to binary, so now that is used instead of
1392 the text comparison.
1394 PH/20 There was another similar case to PH/19, when a complete host name was
1395 given in a host list; looking up its IP address could give an abbreviated
1396 form, whereas the current host's name might or might not be abbreviated.
1397 The same fix has been applied.
1403 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1404 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1407 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1408 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1410 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1412 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1413 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1414 Also, for "Auto-Submitted", treat anything other than "no" as
1415 non-personal, in accordance with RFC 3834. (Previously it treated
1416 anything starting "auto-" as non-personal.)
1418 TF/01 The control=submission/name=... option had a problem with syntax
1419 errors if the name included a slash character. The /name= option
1420 now slurps the rest of the string, so it can include any characters
1421 but it must come last in the list of options (after /sender_retain
1424 PH/05 Some modifications to the interface to the fake nameserver for the new
1432 TK/01 Added the "success_on_redirect" address verification option. See
1433 NewStuff for rationale and an example.
1435 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1437 PH/02 Patch to exigrep to allow it to work on syslog lines.
1439 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1440 fread() to skip over the body file's header line, because in Cygwin the
1441 header line is locked and is inaccessible.
1443 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1444 co-exist for some time) to make it clear that it is the Exim ID that is
1445 referenced, not the Message-ID: header line.
1447 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1448 string_format() function, because snprintf() does not exist on all
1451 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1452 ${stat: expansion item.
1454 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1455 protocol synchronization error", to keep the pedants happy.
1457 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1458 well as for IRIX systems, when gcc is being used. See the host.c source
1461 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1463 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1466 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1467 transport and to the smtp transport in LMTP mode.
1469 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1471 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1472 run a filter in a subprocess. This could lead to confusion in subsequent
1473 lookups in the parent process. There should also be a search_tidyup() at
1474 the end of the subprocess.
1476 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1477 only if the host matched helo_try_verify_hosts, which caused the
1478 verification to occur when the EHLO/HELO command was issued. The ACL just
1479 tested the remembered result. Now, if a previous verification attempt has
1480 not happened, "verify = helo" does it there and then.
1482 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1484 TK/03 Fix log output including CR from clamd.
1486 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1487 exist provoked a memory error which could cause a segfault.
1489 PH/15 Installed PCRE 6.2
1491 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1493 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1494 of the problem. Specifically, suggested +O2 rather than +O1 for the
1497 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1499 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1500 record, the port was not correctly logged when the outgoing_port log
1501 selector was set (it logged the transort's default port).
1503 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1504 fallback_hosts, and "hosts" in the smtp transport.
1506 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1507 host errors such as "Connection refused".
1509 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1510 authentication with radiusclient 0.4.9:
1512 - Error returned from rc_read_config was caught wrongly
1513 - Username/password not passed on to radius server due to wrong length.
1515 The presumption is that some radiusclient API changes for 4.51/PH/17
1516 were not taken care of correctly. The code is still untested by me (my
1517 Linux distribution still has 0.3.2 of radiusclient), but it was
1518 contributed by a Radius user.
1520 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1521 expanding the "port" option of the smtp transport.
1523 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1524 while reading a MIME header. Thanks to Tom Hughes for a patch.
1526 PH/24 Include config.h inside local_scan.h so that configuration settings are
1529 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1530 that in an ACL for RCPT (for example), you can examine exactly what was
1533 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1534 commands, but it was not correctly comparing the address with the actual
1535 client host address. Thus, it would show the EHLO address in Received:
1536 header lines when this was not necessary.
1538 PH/27 Added the % operator to ${eval:}.
1540 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1541 it should be ignoring failures (because with -C, for example, it has lost
1542 privilege). It wasn't ignoring creation failures other than "already
1545 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1548 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1549 because an input line was too long, either on its own, or by virtue of
1550 too many continuations, the temporary file was not being removed, and the
1551 return code was incorrect.
1553 PH/31 Missing "BOOL" in function definition in filtertest.c.
1555 PH/32 Applied Sieve patches from the maintainer.
1557 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1559 PH/33 Added "verify = not_blind".
1561 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1562 Local/Makefile (with some defaults set). These are used in built scripts
1563 such as exicyclog, but they have never been used in the exim_install
1564 script (though there are many overriding facilities there). I have
1565 arranged that the exim_install script now takes note of these two
1568 PH/35 Installed configuration files for Dragonfly.
1570 PH/36 When a locally submitted message by a trusted user did not contain a
1571 From: header, and the sender address was obtained from -f or from an SMTP
1572 MAIL command, and the trusted user did not use -F to supply a sender
1573 name, $originator_name was incorrectly used when constructing a From:
1574 header. Furthermore, $originator_name was used for submission mode
1575 messages from external hosts without From: headers in a similar way,
1576 which is clearly wrong.
1578 PH/37 Added control=suppress_local_fixups.
1580 PH/38 When log_selector = +received_sender was set, and the addition of the
1581 sender made the log line's construction buffer exactly full, or one byte
1582 less than full, an overflow happened when the terminating "\n" was
1585 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1586 when the result of a list match is failure because a DNS lookup failed.
1588 PH/40 RM_COMMAND is now used in the building process.
1590 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1591 the "build-* directories that it finds.
1593 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1594 address in a domain literal was a prefix of an interface address.
1596 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1597 when verifying a sender address, unless rewrite_headers is false.
1599 PH/44 Wrote a long comment about why errors_to addresses are verified as
1600 recipients, not senders.
1602 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1603 the ratelimit ACL was added.
1605 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1607 PH/46 Added extra information about PostgreSQL errors to the error string.
1609 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1610 avoiding the need to install special zones in a real server. This is
1611 backwards compatible; if it can't find the fake resolver, it drops back.
1612 Thus, both old and new test suites can be run.
1614 TF/02 Added util/ratelimit.pl
1616 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1617 clock is set back in time.
1619 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1620 Candler <B.Candler@pobox.com>.
1622 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1623 OK for addresses that are the result of redirections.
1625 PH/48 A number of further additions for the benefit of the new test suite,
1626 including a fake gethostbyname() that interfaces to the fake DNS resolver
1629 TF/06 The fix for widen_domains has also been applied to qualify_single and
1630 search_parents which are the other dnslookup options that can cause
1633 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1636 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1638 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1639 service" if it encountered an MX record with an empty target hostname.
1640 The message is now "an MX or SRV record indicated no SMTP service".
1642 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1643 if the DNS of the sending site is misconfigured. This is quite a
1644 common situation. This change restores the behaviour of treating a
1645 helo verification defer as a failure.
1647 PH/51 If self=fail was set on a router, the bounce message did not include the
1648 actual error message.
1654 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1656 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1657 command itself ended in error, the underlying message about the transport
1658 filter timeout was being overwritten with the pipe command error. Now the
1659 underlying error message should be appended to the second error message.
1661 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1663 PH/02 Reduce the amount of output that "make" produces by default. Full output
1664 can still be requested.
1666 PH/03 The warning log line about a condition test deferring for a "warn" verb
1667 was being output only once per connection, rather than after each
1668 occurrence (because it was using the same function as for successful
1669 "warn" verbs). This seems wrong, so I have changed it.
1671 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1672 it should not have, which might have caused a crash in the right
1673 circumstances, but probably never did.
1675 PH/04 Installed a modified version of Tony Finch's patch to make submission
1676 mode fix the return path as well as the Sender: header line, and to
1677 add a /name= option so that you can make the user's friendly name appear
1680 TF/03 Added the control = fakedefer ACL modifier.
1682 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1683 Mark Lowes for thorough testing.
1685 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1687 TK/03 Merged latest SRS patch from Miles Wilton.
1689 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1690 with the definition in sysexits.h (which is #included earlier).
1691 Fortunately, Exim does not actually use EX_OK. The code used to try to
1692 preserve the sysexits.h value, by assumimg that macro definitions were
1693 scanned for macro replacements. I have been disabused of this notion,
1694 so now the code just undefines EX_OK before #including unistd.h.
1696 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1697 in the smtp transport. When a block could not be written in a single
1698 write() function, the timeout was being re-applied to each part-write.
1699 This seems wrong - if the receiver was accepting one byte at a time it
1700 would take for ever. The timeout is now adjusted when this happens. It
1701 doesn't have to be particularly precise.
1703 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1704 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1706 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1707 without a domain if the check to <postmaster@domain> fails.
1709 SC/01 Eximstats: added -xls and the ability to specify output files
1710 (patch written by Frank Heydlauf).
1712 SC/02 Eximstats: use FileHandles for outputing results.
1714 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1716 SC/04 Eximstats: fixed display of large numbers with -nvr option
1718 SC/05 Eximstats: fixed merging of reports with empty tables.
1720 SC/06 Eximstats: added the -include_original_destination flag
1722 SC/07 Eximstats: removed tabs and trailing whitespace.
1724 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1726 TK/06 MBOX spool code: Add real "From " MBOX separator line
1727 so the .eml file is really in mbox format (even though
1728 most programs do not really care). Patch from Alex Miller.
1730 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1731 The latter is generated from $received_to and is only set if the
1732 message has one envelope recipient. SA can use these headers,
1733 obviously out-of-the-box. Patch from Alex Miller.
1735 PH/08 The ${def test on a variable was returning false if the variable's
1736 value was "0", contrary to what the specification has always said!
1737 The result should be true unless the variable is empty.
1739 PH/09 The syntax error of a character other than { following "${if
1740 def:variable_name" (after optional whitespace) was not being diagnosed.
1741 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1742 accidental colon was present, for example, could give incorrect results.
1744 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1745 result is used (not including appendfile, where other changes are about
1748 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1749 This involved changing a lot of size variables from int to off_t. It
1750 should work with maildirs and everything.
1752 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1753 spamd dying while we are connected to it.
1755 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1758 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1759 function for BDB 4.3.
1761 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1763 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1764 the sources for PCRE have been re-arranged and more files are now
1767 PH/15 The code I had for printing potentially long long variables in PH/11
1768 above was not the best (it lost precision). The length of off_t variables
1769 is now inspected at build time, and an appropriate printing format (%ld
1770 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1771 to be "long long int" or "long int". This is needed for the internal
1772 formatting function string_vformat().
1774 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1775 the configuration file to be ":syslog", then the script "guesses" where
1776 the logs files are, rather than using the compiled in default. In our
1777 case the guess is not the same as the compiled default, so the script
1778 suddenly stopped working when I started to use syslog. The patch checks
1779 to see if log_file_path is "". If so, it attempts to read it from exim
1780 with no configuration file to get the compiled in version, before it
1781 falls back to the previous guessing code."
1783 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1784 implementing BATV in an Exim configuration. See NewStuff for the gory
1787 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1788 Makefile that are specific to HP-UX.
1790 PH/18 If the "use_postmaster" option was set for a recipient callout together
1791 with the "random" option, the postmaster address was used as the MAIL
1792 FROM address for the random test, but not for the subsequent recipient
1793 test. It is now used for both.
1795 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1796 patch removes a few documentation additions to RFC 3028, because the
1797 latest draft now contains them. It adds the new en;ascii-case comparator
1798 and a new error check for 8bit text in MIME parts. Comparator and
1799 require names are now matched exactly. I enabled the subaddress
1800 extension, but it is not well tested yet (read: it works for me)."
1802 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1803 rework some of the code of TK/09 above to avoid the hardwired use of
1804 "%lld" and "long long". Replaced the call to snprintf() with a call to
1807 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1808 records point to non-existent hosts", "retry timeout exceeded", and
1809 "retry time not reached for any host after a long failure period".
1811 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1812 experimental DomainKeys support:
1814 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1815 (2) On an error such as an illegally used "control", the wrong name for
1816 the control was given.
1818 These problems did NOT occur unless DomainKeys support was compiled.
1820 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1822 PH/24 Added ${if match_ip condition.
1824 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1825 databases so that it will be absolutely obvious if a crash occurs in the
1826 DB library. This is a regular occurrence (often caused by mis-matched
1829 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1830 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1831 picked up on a user's system that detects such things. There doesn't seem
1832 to be a gcc warning option for this - only an attribute that has to be
1833 put on the function's prototype. It seems that in Fedora Core 4 they have
1834 set this on a number of new functions. No doubt there will be more in due
1837 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1838 specify a transport. However, if an address that was verified by such a
1839 router was the subject of a callout, Exim crashed because it tried to
1840 read the rcpt_include_affixes from the non-existent transport. Now it
1841 just assumes that the setting of that option is false. This bug was
1842 introduced by 4.51/PH/31.
1844 PH/28 Changed -d+all to exclude +memory, because that information is very
1845 rarely of interest, but it makes the output a lot bigger. People tend to
1846 do -d+all out of habit.
1848 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1849 code in os-type was giving problems when libc.so lives in lib64, like on
1852 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1853 aren't the modern standard, and it seems that some systems' include files
1854 don't always have them. Exim was already checking for some of the newer
1855 ones like T_AAAA, and defining it itself. I've added checks for all the
1856 record types that Exim uses.
1858 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1859 not automatically generating a new one, as it is supposed to. This
1860 prevented TLS from working. If the file did exist, but contained invalid
1861 data, a new version was generated, as expected. It was only the case of a
1862 non-existent file that was broken.
1864 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1865 with a change in libdomainkeys > 0.64.
1867 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1868 from DNS. If the selector record carries the flag, it now has
1869 precedence over the domain-wide flag.
1871 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1873 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1874 the use of a port name that isn't defined in /etc/services) occurred, the
1875 message was deferred as in a normal delivery, and thus remained on the
1876 spool, instead of being failed because of the mua_wrapper setting. This
1877 is now fixed, and I tidied up some of the mua_wrapper messages at the
1880 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1881 the messages in a hash of arrays rather than using individual hashes.
1882 This is a bit cleaner and results in dramatic memory savings, albeit
1883 at a slight CPU cost.
1885 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1886 as requested by Marc Sherman.
1888 SC/10 Eximstats: added histograms for user specified patterns as requested
1891 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1893 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1894 fopen() in the content-scanning modules that did not already have it.
1900 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1901 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1903 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1905 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1907 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1908 to test Sieve filters that use "vacation".
1910 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1911 that changes the way the GnuTLS parameters are stored in the cache file.
1912 The new format can be generated externally. For backward compatibility,
1913 if the data in the cache doesn't make sense, Exim assumes it has read an
1914 old-format file, and it generates new data and writes a new file. This
1915 means that you can't go back to an older release without removing the
1918 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1919 work if there are any delivery delays because "one_time" forces the
1920 parent to be marked "delivered", so its unseen clone is never tried
1921 again. For this reason, Exim now forbids the simultaneous setting of
1924 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1925 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1926 entry if you want to know the details. The fix, however, neglected to
1927 consider the case when local delivery batching is involved. The test for
1928 "previously delivered" was not happening when checking to see if an
1929 address could be batched with a previous (undelivered) one; under
1930 certain circumstances this could lead to multiple deliveries to the same
1933 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1934 in its include files, and this causes problems building Exim.
1936 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1937 header_syntax) but Exim was just ignoring anything given after a slash.
1938 In particular, this caused confusion with an attempt to use "verify =
1939 reverse_host_lookup/defer_ok". An error is now given when options are
1940 supplied for verify items that do not have them. (Maybe reverse_host_
1941 lookup should have a defer_ok option, but that's a different point.)
1943 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1944 defined by RFC 821) to 2048, because there were problems with some AUTH
1945 commands, and RFC 1869 says the size should be increased for extended
1946 SMTP commands that take arguments.
1948 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1951 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1952 "unknown" error; now it says that the functionality isn't in the binary.
1954 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1955 an address' error message when a string expansion fails (syntax or
1956 whatever). Otherwise the password may appear in the log. Following change
1957 PH/42 below, there is no longer a chance of it appearing in a bounce
1960 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1962 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1963 of its addresses was ever tried. (Bugzilla bug #2.)
1965 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1966 the result incorrectly in the debug output. (It correctly added a newline
1967 to what was transported.)
1969 TF/01 Added $received_time.
1971 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1972 commented out examples of how to interface to a virus scanner and to
1973 SpamAssassin. Also added commented examples of av_scanner and
1974 spamd_address settings.
1976 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1977 and controls are allowed in which ACLs. There were a couple of minor
1978 errors. Some of the entries in the conditions table (which is a table of
1979 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1980 negation of where the condition IS allowed.
1982 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1984 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1985 header file does not have a version number, so I've had to invent a new
1986 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1987 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1988 radiusclient), but it was contributed by a Radius user.
1990 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1991 files or directories, for OpenSSL.
1993 PH/19 When an Exim process that is running as root has to create an Exim log
1994 file, it does so in a subprocess that runs as exim:exim so as to get the
1995 ownership right at creation (otherwise, other Exim processes might see
1996 the file with the wrong ownership). There was no test for failure of this
1997 fork() call, which would lead to the process getting stuck as it waited
1998 for a non-existent subprocess. Forks do occasionally fail when resources
1999 run out. I reviewed all the other calls to fork(); they all seem to check
2002 PH/20 When checking for unexpected SMTP input at connect time (before writing
2003 the banner), Exim was not dealing correctly with a non-positive return
2004 from the read() function. If the client had disconnected by this time,
2005 the result was a log entry for a synchronization error with an empty
2006 string after "input=" when read() returned zero. If read() returned -1
2007 (an event I could not check), uninitialized data bytes were printed.
2008 There were reports of junk text (parts of files, etc) appearing after
2011 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2013 PH/22 Added support for macro redefinition, and (re)definition in between
2014 driver and ACL definitions.
2016 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2017 forgetting to use the resulting value; it was using the unexpanded value.
2019 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2020 hadn't been configured. The fix is from Juergen Kreileder, who
2021 understands it better than I do:
2023 "Here's what I see happening with three configured cyrus_sasl
2024 authenticators configured (plain, login, cram-md5):
2026 On startup auth_cyrus_sasl_init() gets called for each of these.
2027 This means three calls to sasl_listmech() without a specified mech_list.
2028 => SASL tests which mechs of all available mechs actually work
2029 => three warnings about OTP not working
2030 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2032 With the patch, sasl_listmech() also gets called three times. But now
2033 SASL's mech_list option is set to the server_mech specified in the the
2034 authenticator. Or in other words, the answer from sasl_listmech()
2035 gets limited to just the mech you're testing for (which is different
2037 => the return list contains just 'plain' or 'login', 'cram-md5' or
2038 nothing depending on the value of ob->server_mech.
2040 I've just tested the patch: Authentication still works fine,
2041 unavailable mechs specified in the exim configuration are still
2042 caught, and the auth.log warnings about OTP are gone."
2044 PH/25 When debugging is enabled, the contents of the command line are added
2045 to the debugging output, even when log_selector=+arguments is not
2048 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2049 answer is "GNU", and only if the return is "GNU/something" is the answer
2052 PH/27 $acl_verify_message is now set immediately after the failure of a
2053 verification in an ACL, and so is available in subsequent modifiers. In
2054 particular, the message can be preserved by coding like this:
2056 warn !verify = sender
2057 set acl_m0 = $acl_verify_message
2059 Previously, $acl_verify_message was set only while expanding "message"
2060 and "log_message" when a very denied access.
2062 PH/28 Modified OS/os.c-Linux with
2064 -#ifndef OS_LOAD_AVERAGE
2065 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2067 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2068 nomenclature these days.)
2070 PH/29 Installed patch from the Sieve maintainer that adds the options
2071 sieve_useraddress and sieve_subaddress to the redirect router.
2073 PH/30 In these circumstances:
2074 . Two addresses routed to the same list of hosts;
2075 . First host does not offer TLS;
2076 . First host accepts first address;
2077 . First host gives temporary error to second address;
2078 . Second host offers TLS and a TLS session is established;
2079 . Second host accepts second address.
2080 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2081 and peerdn, if requested) that were in fact used only for the second
2084 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2085 attention to any local part prefix or suffix that was matched by the
2086 router that accepted the address. It now behaves in the same way as it
2087 does for delivery: the affixes are removed from the local part unless
2088 rcpt_include_affixes is set on the transport.
2090 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2091 timeout during the DATA phase of an incoming message.
2093 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2094 applied a patch sanctioned by the Sieve maintainer.
2096 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2097 the uid or gid is negative. A case of a negative gid caused this to be
2098 noticed. The fix allows for either to be negative.
2100 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2101 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2104 PH/36 The change PH/12 above was broken. Fixed it.
2106 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2107 the grounds that routing the same address twice would always produce the
2108 same answer. This might have been true once, but it is certainly no
2109 longer true now. Routing a child address may depend on the previous
2110 routing that produced that child. Some complicated redirection strategies
2111 went wrong when messages had multiple recipients, and made Exim's
2112 behaviour dependent on the order in which the addresses were given.
2114 I have moved the duplicate checking until after the routing is complete.
2115 Exim scans the addresses that are assigned to local and remote
2116 transports, and removes any duplicates. This means that more work will be
2117 done, as duplicates will always all be routed, but duplicates are
2118 presumably rare, so I don't expect this is of any significance.
2120 For deliveries to pipes, files, and autoreplies, the duplicate checking
2121 still happens during the routing process, since they are not going to be
2124 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2125 It corrects a timeout issue with spamd. This is Ian's comment: "The
2126 background is that sometimes spamd either never reads data from a
2127 connection it has accepted, or it never writes response data. The exiscan
2128 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2129 blindly assumes that writes won't block so it may never time out."
2131 PH/39 Allow G after quota size as well as K and M.
2133 PH/40 The value set for $authenticated_id in an authenticator may not contain
2134 binary zeroes or newlines because the value is written to log lines and
2135 to spool files. There was no check on this. Now the value is run through
2136 the string_printing() function so that such characters are converted to
2137 printable escape sequences.
2139 PH/41 $message_linecount is a new variable that contains the total number of
2140 lines in the message. Compare $body_linecount, which is the count for the
2143 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2144 bounce and delay warning messages, except in certain special cases, which
2147 (a) An SMTP error message from a remote host;
2148 (b) A message specified in a :fail: redirection;
2149 (c) A message specified in a "fail" command in a system filter;
2150 (d) A message specified in a FAIL return from the queryprogram router;
2151 (e) A message specified by the cannot_route_message router option.
2153 In these cases only, Exim does include the error details in bounce and
2154 warning messages. There are also a few cases where bland messages such
2155 as "unrouteable address" or "local delivery error" are given.
2157 PH/43 $value is now also set for the "else" part of a ${run expansion.
2159 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2160 being worked on, but at least Exim now implements the latest version to
2163 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2164 process to complete was treated as a delivery failure, a timeout while
2165 writing the message to the pipe was logged, but erroneously treated as a
2166 successful delivery. Such timeouts include transport filter timeouts. For
2167 consistency with the overall process timeout, these timeouts are now
2168 treated as errors, giving rise to delivery failures by default. However,
2169 there is now a new Boolean option for the pipe transport called
2170 timeout_defer, which, if set TRUE, converts the failures into defers for
2171 both kinds of timeout. A transport filter timeout is now identified in
2174 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2175 systems where "make" and "gmake" are different, calling "gmake" at top
2176 level broke things. I've arranged for the value of $(MAKE) to be passed
2177 from the Makefile to this script so that it can call the same version of
2181 A note about Exim versions 4.44 and 4.50
2182 ----------------------------------------
2184 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2185 changes of various kinds. As a consequence, a big documentation update was
2186 needed. This delayed the release for rather longer than seemed good, especially
2187 in the light of a couple of (minor) security issues. Therefore, the changes
2188 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2189 release. So 4.44 and 4.50 are in effect two different branches that both start
2192 I have left the 4.50 change log unchanged; it contains all the changes since
2193 4.43. The change log for 4.44 is below; many of its items are identical to
2194 those for 4.50. This seems to be the most sensible way to preserve the
2195 historical information.
2201 1. Minor wording change to the doc/README.SIEVE file.
2203 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2204 computation of the current number of files was incorrect.
2206 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2207 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2210 4. Give more explanation in the error message when the command for a transport
2211 filter fails to execute.
2213 5. There are several places where Exim runs a non-Exim command in a
2214 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2215 was being done only for the command run by the queryprogram router. It is
2216 now done for all such subprocesses. The other cases are: ${run, transport
2217 filters, and the commands run by the lmtp and pipe transports.
2219 6. Added CONFIGURE_GROUP build-time option.
2221 7. Some older OS have a limit of 256 on the maximum number of file
2222 descriptors. Exim was using setrlimit() to set 1000 as a large value
2223 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2224 systems. I've change it so that if it can't get 1000, it tries for 256.
2226 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2227 was an oversight, and furthermore, ever since the addition of extra
2228 controls (e.g. 4.43/32), the checks on when to allow different forms of
2229 "control" were broken. There should now be diagnostics for all cases when a
2230 control that does not make sense is encountered.
2232 9. Added the /retain_sender option to "control=submission".
2234 10. $recipients is now available in the predata ACL (oversight).
2236 11. Tidy the search cache before the fork to do a delivery from a message
2237 received from the command line. Otherwise the child will trigger a lookup
2238 failure and thereby defer the delivery if it tries to use (for example) a
2239 cached ldap connection that the parent has called unbind on.
2241 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2242 of $address_data from the recipient verification was clobbered by the
2243 sender verification.
2245 13. The value of address_data from a sender verification is now available in
2246 $sender_address_data in subsequent conditions in the ACL statement.
2248 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2250 15. Added a new option "connect=<time>" to callout options, to set a different
2253 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2254 was its contents. (It was OK if the option was not defined at all.)
2256 17. A "Completed" log line is now written for messages that are removed from
2257 the spool by the -Mrm option.
2259 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2260 information about exactly what failed.
2262 19. Added -dd to debug only the daemon process.
2264 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2265 handles timeouts, both on the server side and network timeouts. Renamed the
2266 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2268 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2269 It is now set to "smtps".
2271 22. $host_address is now set to the target address during the checking of
2272 ignore_target_hosts.
2274 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2275 being passed; this would have caused $sender_host_name to have been used if
2276 matching the list had actually called for a host name (not very likely,
2277 since this list is usually IP addresses). A host name is now passed as
2280 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2281 code that specifies a non-restarting handler (typically sigaction() in
2282 modern systems) in an attempt to fix a rare and obscure crash bug.
2284 25. Narrowed the window for a race in the daemon that could cause it to ignore
2285 SIGCHLD signals. This is not a major problem, because they are used only to
2286 wake it up if nothing else does.
2288 26. A malformed maildirsize file could cause Exim to calculate negative values
2289 for the mailbox size or file count. Odd effects could occur as a result.
2290 The maildirsize information is now recalculated if the size or filecount
2293 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2294 support for a long time. Removed HAVE_SYS_VFS_H.
2296 28. Installed the latest version of exipick from John Jetmore.
2298 29. In an address list, if the pattern was not a regular expression, an empty
2299 subject address (from a bounce message) matched only if the pattern was an
2300 empty string. Non-empty patterns were not even tested. This was the wrong
2301 because it is perfectly reasonable to use an empty address as part of a
2302 database query. An empty address is now tested by patterns that are
2303 lookups. However, all the other forms of pattern expect the subject to
2304 contain a local part and a domain, and therefore, for them, an empty
2305 address still always fails if the pattern is not itself empty.
2307 30. Exim went into a mad DNS loop when attempting to do a callout where the
2308 host was specified on an smtp transport, and looking it up yielded more
2309 than one IP address.
2311 31. Re-factored the code for checking spool and log partition space into a
2312 function that finds that data and another that does the check. The former
2313 is then used to implement four new variables: $spool_space, $log_space,
2314 $spool_inodes, and $log_inodes.
2316 32. The RFC2047 encoding function was originally intended for short strings
2317 such as real names; it was not keeping to the 75-character limit for
2318 encoded words that the RFC imposes. It now respects the limit, and
2319 generates multiple encoded words if necessary. To be on the safe side, I
2320 have increased the buffer size for the ${rfc2047: expansion operator from
2323 33. It is now permitted to omit both strings after an "if" condition; if the
2324 condition is true, the result is "true". As before, when the second string
2325 is omitted, a false condition yields an empty string. This makes it less
2326 cumbersome to write custom ACL and router conditions.
2328 34. Failure to deliver a bounce message always caused it to be frozen, even if
2329 there was an errors_to setting on the router. The errors_to setting is now
2332 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2333 canonical form (fully expanded) before being placed in
2334 $sender_host_address.
2336 36. The table in the code that translates DNS record types into text (T_A to
2337 "A" for instance) was missing entries for NS and CNAME. It is just possible
2338 that this could have caused confusion if both these types were looked up
2339 for the same domain, because the text type is used as part of Exim's
2340 per-process caching. But the chance of anyone hitting this buglet seems
2343 37. The dnsdb lookup has been extended in a number of ways.
2345 (1) There is a new type, "zns", which walks up the domain tree until it
2346 finds some nameserver records. It should be used with care.
2348 (2) There is a new type, "mxh", which is like "mx" except that it returns
2349 just the host names, not the priorities.
2351 (3) It is now possible to give a list of domains (or IP addresses) to be
2352 looked up. The behaviour when one of the lookups defers can be
2353 controlled by a keyword.
2355 (4) It is now possible to specify the separator character for use when
2356 multiple records are returned.
2358 38. The dnslists ACL condition has been extended: it is now possible to supply
2359 a list of IP addresses and/or domains to be looked up in a particular DNS
2362 39. Added log_selector=+queue_time_overall.
2364 40. When running the queue in the test harness, wait just a tad after forking a
2365 delivery process, to get repeatability of debugging output.
2367 41. Include certificate and key file names in error message when GnuTLS fails
2368 to set them up, because the GnuTLS error message doesn't include the name
2369 of the failing file when there is a problem reading it.
2371 42. Allow both -bf and -bF in the same test run.
2373 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2375 44. The "Exiscan patch" is now merged into the mainline Exim source.
2377 45. Sometimes the final signoff response after QUIT could fail to get
2378 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2379 < 0 before doing a fflush(). This bug looks as though it goes back to the
2380 introduction of TLS in release 3.20, but "sometimes" must have been rare
2381 because the tests only now provoked it.
2383 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2384 (this can affect the format of dates).
2386 47. exim_tidydb, when checking for the continued existence of a message for
2387 which it has found a message-specific retry record, was not finding
2388 messages that were in split spool directories. Consequently, it was
2389 deleting retry records that should have stayed in existence.
2391 48. Steve fixed some bugs in eximstats.
2393 49. The SPA authentication driver was not abandoning authentication and moving
2394 on to the next authenticator when an expansion was forced to fail,
2395 contradicting the general specification for all authenticators. Instead it
2396 was generating a temporary error. It now behaves as specified.
2398 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2399 (the order specifies the preference for clients). The order is now AES256,
2400 AES128, 3DES, ARCFOUR128.
2402 51. Small patch to Sieve code - explicitly set From: when generating an
2405 52. Exim crashed if a remote delivery caused a very long error message to be
2406 recorded - for instance if somebody sent an entire SpamAssassin report back
2407 as a large number of 550 error lines. This bug was coincidentally fixed by
2408 increasing the size of one of Exim's internal buffers (big_buffer) that
2409 happened as part of the Exiscan merge. However, to be on the safe side, I
2410 have made the code more robust (and fixed the comments that describe what
2413 53. Now that there can be additional text after "Completed" in log lines (if
2414 the queue_time_overall log selector is set), a one-byte patch to exigrep
2415 was needed to allow it to recognize "Completed" as not the last thing in
2418 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2419 patch that reportedly fixes this has been added. I am not expert enough to
2420 create a test for it. This is what the patch creator wrote:
2422 "I found a little strange behaviour of ldap code when working with
2423 Windows 2003 AD Domain, where users was placed in more than one
2424 Organization Units. When I tried to give exim partial DN, the exit code
2425 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2426 But simultaneously result of request was absolutely normal ldap result,
2427 so I produce this patch..."
2429 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2430 REFERENCE, so I have modified the code to exclude the patch when that macro
2433 55. Some experimental protocols are using DNS PTR records for new purposes. The
2434 keys for these records are domain names, not reversed IP addresses. The
2435 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2436 leaves it alone. Component reversal etc. now happens only for IP addresses.
2439 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2441 57. Double the size of the debug message buffer (to 2048) so that more of very
2442 long debug lines gets shown.
2444 58. The exicyclog utility now does better if the number of log files to keep
2445 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2447 59. Two changes related to the smtp_active_hostname option:
2449 (1) $smtp_active_hostname is now available as a variable.
2450 (2) The default for smtp_banner uses $smtp_active_hostname instead
2451 of $primary_hostname.
2453 60. The host_aton() function is supposed to be passed a string that is known
2454 to be a valid IP address. However, in the case of IPv6 addresses, it was
2455 not checking this. This is a hostage to fortune. Exim now panics and dies
2456 if the condition is not met. A case was found where this could be provoked
2457 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2458 components; fortuitously, this particular loophole had already been fixed
2459 by change 4.50/55 above.
2461 If there are any other similar loopholes, the new check in host_aton()
2462 itself should stop them being exploited. The report I received stated that
2463 data on the command line could provoke the exploit when Exim was running as
2464 exim, but did not say which command line option was involved. All I could
2465 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2466 running as the user.
2469 61. There was a buffer overflow vulnerability in the SPA authentication code
2470 (which came originally from the Samba project). I have added a test to the
2471 spa_base64_to_bits() function which I hope fixes it.
2474 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2475 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2477 63. The daemon start-up calls getloadavg() while still root for those OS that
2478 need the first call to be done as root, but it missed one case: when
2479 deliver_queue_load_max is set with deliver_drop_privilege. This is
2480 necessary for the benefit of the queue runner, because there is no re-exec
2481 when deliver_drop_privilege is set.
2483 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2484 This has been fixed.
2486 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2487 was in use, was not putting the data itself into the right store pool;
2488 consequently, it could be overwritten for a subsequent message in the same
2489 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2492 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2494 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2495 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2496 the function were treating the return as a boolean value, which happened to
2497 work because 0=false and not-0=true, but is not correct code.
2499 68. The host_aton() function was not handling scoped IPv6 addresses (those
2500 with, for example, "%eth0" on the end) correctly.
2502 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2503 negated items (that is, ~something) in unsigned ints. Some compilers
2504 apparently mutter when there is no cast.
2506 70. If an address verification called from an ACL failed, and did not produce a
2507 user-specific message (i.e. there was only a "system" message), nothing was
2508 put in $acl_verify_message. In this situation, it now puts the system
2511 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2512 session; change 4.31/43 added the unwanted input to the log line - except
2513 that it did not do this in the start of session case. It now does.
2515 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2516 This is wrong and can cause the other end to generate a synchronization
2517 error if it is another Exim or anything else that does the synchronization
2518 check. A QUIT command is no longer sent after a timeout.
2520 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2521 during host lookups.
2523 74. The defer_ok option of callout verification was not working if it was used
2524 when verifying addresses in header lines, that is, for this case:
2526 verify = header_sender/callout=defer_ok
2528 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2529 those file descriptors could be used for SMTP connections. If anything
2530 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2531 could be sent to the SMTP client, causing chaos. The daemon now opens
2532 stdin, stdout, and stderr to /dev/null when it puts itself into the
2535 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2536 log by default. The user can override this with suitable Perl magic.
2538 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2539 the log message when discard triggers, was not working for the DATA ACL or
2540 for the non-SMTP ACL.
2542 78. Error message wording change in sieve.c.
2544 79. If smtp_accept_max_per_host was set, the number of connections could be
2545 restricted to fewer than expected, because the daemon was trying to set up
2546 a new connection before checking whether the processes handling previous
2547 connections had finished. The check for completed processes is now done
2548 earlier. On busy systems, this bug wouldn't be noticed because something
2549 else would have woken the daemon, and it would have reaped the completed
2552 80. If a message was submitted locally by a user whose login name contained one
2553 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2554 It caused a spool format error. I have fixed the spool reading code. A
2555 related problem was that the "from" clause in the Received: line became
2556 illegal because of the space(s). It is now covered by ${quote_local_part.
2558 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2561 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2563 83. Patch from Sieve maintainer to fix unterminated string problem in
2564 "vacation" handling.
2566 84. Some minor changes to the Linux configuration files to help with other
2567 OS variants using glibc.
2569 85. One more patch for Sieve to update vacation handling to latest spec.
2572 ----------------------------------------------------
2573 See the note above about the 4.44 and 4.50 releases.
2574 ----------------------------------------------------
2580 1. Change 4.43/35 introduced a bug that caused file counts to be
2581 incorrectly computed when quota_filecount was set in an appendfile
2584 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2585 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2588 3. Give more explanation in the error message when the command for a transport
2589 filter fails to execute.
2591 4. There are several places where Exim runs a non-Exim command in a
2592 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2593 was being done only for the command run by the queryprogram router. It is
2594 now done for all such subprocesses. The other cases are: ${run, transport
2595 filters, and the commands run by the lmtp and pipe transports.
2597 5. Some older OS have a limit of 256 on the maximum number of file
2598 descriptors. Exim was using setrlimit() to set 1000 as a large value
2599 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2600 systems. I've change it so that if it can't get 1000, it tries for 256.
2602 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2603 was an oversight, and furthermore, ever since the addition of extra
2604 controls (e.g. 4.43/32), the checks on when to allow different forms of
2605 "control" were broken. There should now be diagnostics for all cases when a
2606 control that does not make sense is encountered.
2608 7. $recipients is now available in the predata ACL (oversight).
2610 8. Tidy the search cache before the fork to do a delivery from a message
2611 received from the command line. Otherwise the child will trigger a lookup
2612 failure and thereby defer the delivery if it tries to use (for example) a
2613 cached ldap connection that the parent has called unbind on.
2615 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2616 of $address_data from the recipient verification was clobbered by the
2617 sender verification.
2619 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2620 was its contents. (It was OK if the option was not defined at all.)
2622 11. A "Completed" log line is now written for messages that are removed from
2623 the spool by the -Mrm option.
2625 12. $host_address is now set to the target address during the checking of
2626 ignore_target_hosts.
2628 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2629 being passed; this would have caused $sender_host_name to have been used if
2630 matching the list had actually called for a host name (not very likely,
2631 since this list is usually IP addresses). A host name is now passed as
2634 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2635 code that specifies a non-restarting handler (typically sigaction() in
2636 modern systems) in an attempt to fix a rare and obscure crash bug.
2638 15. Narrowed the window for a race in the daemon that could cause it to ignore
2639 SIGCHLD signals. This is not a major problem, because they are used only to
2640 wake it up if nothing else does.
2642 16. A malformed maildirsize file could cause Exim to calculate negative values
2643 for the mailbox size or file count. Odd effects could occur as a result.
2644 The maildirsize information is now recalculated if the size or filecount
2647 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2648 support for a long time. Removed HAVE_SYS_VFS_H.
2650 18. Updated exipick to current release from John Jetmore.
2652 19. Allow an empty sender to be matched against a lookup in an address list.
2653 Previously the only cases considered were a regular expression, or an
2656 20. Exim went into a mad DNS lookup loop when doing a callout where the
2657 host was specified on the transport, if the DNS lookup yielded more than
2660 21. The RFC2047 encoding function was originally intended for short strings
2661 such as real names; it was not keeping to the 75-character limit for
2662 encoded words that the RFC imposes. It now respects the limit, and
2663 generates multiple encoded words if necessary. To be on the safe side, I
2664 have increased the buffer size for the ${rfc2047: expansion operator from
2667 22. Failure to deliver a bounce message always caused it to be frozen, even if
2668 there was an errors_to setting on the router. The errors_to setting is now
2671 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2672 canonical form (fully expanded) before being placed in
2673 $sender_host_address.
2675 24. Updated eximstats to version 1.33
2677 25. Include certificate and key file names in error message when GnuTLS fails
2678 to set them up, because the GnuTLS error message doesn't include the name
2679 of the failing file when there is a problem reading it.
2681 26. Expand error message when OpenSSL has problems setting up cert/key files.
2684 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2685 (this can affect the format of dates).
2687 28. exim_tidydb, when checking for the continued existence of a message for
2688 which it has found a message-specific retry record, was not finding
2689 messages that were in split spool directories. Consequently, it was
2690 deleting retry records that should have stayed in existence.
2692 29. eximstats updated to version 1.35
2693 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2694 1.35 - bugfix such that pie charts by volume are generated correctly
2696 30. The SPA authentication driver was not abandoning authentication and moving
2697 on to the next authenticator when an expansion was forced to fail,
2698 contradicting the general specification for all authenticators. Instead it
2699 was generating a temporary error. It now behaves as specified.
2701 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2702 (the order specifies the preference for clients). The order is now AES256,
2703 AES128, 3DES, ARCFOUR128.
2705 31. Small patch to Sieve code - explicitly set From: when generating an
2708 32. Exim crashed if a remote delivery caused a very long error message to be
2709 recorded - for instance if somebody sent an entire SpamAssassin report back
2710 as a large number of 550 error lines. This bug was coincidentally fixed by
2711 increasing the size of one of Exim's internal buffers (big_buffer) that
2712 happened as part of the Exiscan merge. However, to be on the safe side, I
2713 have made the code more robust (and fixed the comments that describe what
2716 33. Some experimental protocols are using DNS PTR records for new purposes. The
2717 keys for these records are domain names, not reversed IP addresses. The
2718 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2719 leaves it alone. Component reversal etc. now happens only for IP addresses.
2722 34. The host_aton() function is supposed to be passed a string that is known
2723 to be a valid IP address. However, in the case of IPv6 addresses, it was
2724 not checking this. This is a hostage to fortune. Exim now panics and dies
2725 if the condition is not met. A case was found where this could be provoked
2726 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2727 components; fortuitously, this particular loophole had already been fixed
2728 by change 4.50/55 or 4.44/33 above.
2730 If there are any other similar loopholes, the new check in host_aton()
2731 itself should stop them being exploited. The report I received stated that
2732 data on the command line could provoke the exploit when Exim was running as
2733 exim, but did not say which command line option was involved. All I could
2734 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2735 running as the user.
2738 35. There was a buffer overflow vulnerability in the SPA authentication code
2739 (which came originally from the Samba project). I have added a test to the
2740 spa_base64_to_bits() function which I hope fixes it.
2743 36. The daemon start-up calls getloadavg() while still root for those OS that
2744 need the first call to be done as root, but it missed one case: when
2745 deliver_queue_load_max is set with deliver_drop_privilege. This is
2746 necessary for the benefit of the queue runner, because there is no re-exec
2747 when deliver_drop_privilege is set.
2749 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2750 was in use, was not putting the data itself into the right store pool;
2751 consequently, it could be overwritten for a subsequent message in the same
2752 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2755 38. Sometimes the final signoff response after QUIT could fail to get
2756 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2757 < 0 before doing a fflush(). This bug looks as though it goes back to the
2758 introduction of TLS in release 3.20, but "sometimes" must have been rare
2759 because the tests only now provoked it.
2765 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2766 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2767 Now it returns an integer. A number of calls were still expecting a T/F
2768 return. Fortuitously, in all cases, the tests worked in OK situations,
2769 which is the norm. However, things would have gone wrong on any write
2770 failures on the smtp file descriptor. This function is used when sending
2771 messages over SMTP and also when doing verify callouts.
2773 2. When Exim is called to do synchronous delivery of a locally submitted
2774 message (the -odf or -odi options), it no longer closes stderr before doing
2777 3. Implemented the mua_wrapper option.
2779 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2781 5. Implemented the functions header_remove(), header_testname(),
2782 header_add_at_position(), and receive_remove_recipient(), and exported them
2785 6. If an ACL "warn" statement specified the addition of headers, Exim already
2786 inserted X-ACL-Warn: at the start if there was no header name. However, it
2787 was not making this test for the second and subsequent header lines if
2788 there were newlines in the string. This meant that an invalid header could
2789 be inserted if Exim was badly configured.
2791 7. Allow an ACL "warn" statement to add header lines at the start or after all
2792 the Received: headers, as well as at the end.
2794 8. Added the rcpt_4xx retry error code.
2796 9. Added postmaster_mailfrom=xxx to callout verification option.
2798 10. Added mailfrom=xxxx to the callout verification option, for verify=
2801 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2802 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2804 12. Inserted some casts to stop certain compilers warning when using pointer
2805 differences as field lengths or precisions in printf-type calls (mostly
2806 affecting debugging statements).
2808 13. Added optional readline() support for -be (dynamically loaded).
2810 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2811 same clock tick as a message's arrival, so that its received time was the
2812 same as the "first fail" time on the retry record, and that message
2813 remained on the queue past the ultimate address timeout, every queue runner
2814 would try a delivery (because it was past the ultimate address timeout) but
2815 after another failure, the ultimate address timeout, which should have then
2816 bounced the address, did not kick in. This was a "< instead of <=" error;
2817 in most cases the first failure would have been in the next clock tick
2818 after the received time, and all would be well.
2820 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2821 being recognized when the domain list was tested by the match_domain
2822 condition in an expansion string.
2824 16. Added the ${str2b64: operator.
2826 17. Exim was always calling setrlimit() to set a large limit for the number of
2827 processes, without checking whether the existing limit was already
2828 adequate. (It did check for the limit on file descriptors.) Furthermore,
2829 errors from getrlimit() and setrlimit() were being ignored. Now they are
2830 logged to the main and panic logs, but Exim does carry on, to try to do its
2831 job under whatever limits there are.
2833 18. Imported PCRE 5.0.
2835 19. Trivial typo in log message " temporarily refused connection" (the leading
2838 20. If the log selector return_path_on_delivery was set and an address was
2839 redirected to /dev/null, the delivery process crashed because it assumed
2840 that a return path would always be set for a "successful" delivery. In this
2841 case, the whole delivery is bypassed as an optimization, and therefore no
2844 21. Internal re-arrangement: the function for sending a challenge and reading
2845 a response while authentication was assuming a zero-terminated challenge
2846 string. It's now changed to take a pointer and a length, to allow for
2847 binary data in such strings.
2849 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2851 23. Exim was not respecting finduser_retries when seeking the login of the
2852 uid under which it was called; it was always trying 10 times. (The default
2853 setting of finduser_retries is zero.) Also, it was sleeping after the final
2854 failure, which is pointless.
2856 24. Implemented tls_on_connect_ports.
2858 25. Implemented acl_smtp_predata.
2860 26. If the domain in control=submission is set empty, Exim assumes that the
2861 authenticated id is a complete email address when it generates From: or
2862 Sender: header lines.
2864 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2865 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2866 chown and chgrp in /bin and hostname in /usr/bin.
2868 28. Exim was keeping the "process log" file open after each use, just as it
2869 does for the main log. This opens the possibility of it remaining open for
2870 long periods when the USR1 signal hits a daemon. Occasional processlog
2871 errors were reported, that could have been caused by this. Anyway, it seems
2872 much more sensible not to leave this file open at all, so that is what now
2875 29. The long-running daemon process does not normally write to the log once it
2876 has entered its main loop, and it closes the log before doing so. This is
2877 so that log files can straightforwardly be renamed and moved. However,
2878 there are a couple of unusual error situations where the daemon does write
2879 log entries, and I had neglected to close the log afterwards.
2881 30. The text of an SMTP error response that was received during a remote
2882 delivery was being truncated at 512 bytes. This is too short for some of
2883 the long messages that one sometimes sees. I've increased the limit to
2886 31. It is now possible to make retry rules that apply only when a message has a
2887 specific sender, in particular, an empty sender.
2889 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2890 it possible to be selective about when SMTP synchronization is enforced.
2892 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2894 32. Implemented hosts_connection_nolog.
2896 33. Added an ACL for QUIT.
2898 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2901 35. Added mailbox_size and mailbox_filecount to appendfile.
2903 36. Added control = no_multiline_responses to ACLs.
2905 37. There was a bug in the logic of the code that waits for the clock to tick
2906 in the case where the clock went backwards by a substantial amount such
2907 that the microsecond fraction of "now" was more than the microsecond
2908 fraction of "then" (but the whole seconds number was less).
2910 38. Added support for the libradius Radius client library this is found on
2911 FreeBSD (previously only the radiusclient library was supported).
2917 1. When certain lookups returned multiple values in the form name=value, the
2918 quoting of the values was not always being done properly. Specifically:
2919 (a) If the value started with a double quote, but contained no whitespace,
2921 (b) If the value contained whitespace other than a space character (i.e.
2922 tabs or newlines or carriage returns) it was not quoted.
2923 This fix has been applied to the mysql and pgsql lookups by writing a
2924 separate quoting function and calling it from the lookup code. The fix
2925 should probably also be applied to nisplus, ibase and oracle lookups, but
2926 since I cannot test any of those, I have not disturbed their existing code.
2928 2. A hit in the callout cache for a specific address caused a log line with no
2929 reason for rejecting RCPT. Now it says "Previous (cached) callout
2930 verification failure".
2932 3. There was an off-by-one bug in the queryprogram router. An over-long
2933 return line was truncated at 256 instead of 255 characters, thereby
2934 overflowing its buffer with the terminating zero. As well as fixing this, I
2935 have increased the buffer size to 1024 (and made a note to document this).
2937 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2938 when Exim is waiting for an SMTP response from a remote server, Exim
2939 restarts its select() call on the socket, thereby resetting its timeout.
2940 This is not a problem when such interrupts are rare. Somebody set up a cron
2941 job to run exiwhat every 2 minutes, which is less than the normal select()
2942 timeout (5 or 10 minutes). This meant that the select() timeout never
2943 kicked in because it was always reset. I have fixed this by comparing the
2944 time when an interrupt arrives with the time at the start of the first call
2945 to select(). If more time than the timeout has elapsed, the interrupt is
2946 treated as a timeout.
2948 5. Some internal re-factoring in preparation for the addition of Sieve
2949 extensions (by MH). In particular, the "personal" test is moved to a
2950 separate function, and given an option for scanning Cc: and Bcc: (which is
2951 not set for Exim filters).
2953 6. When Exim created an email address using the login of the caller as the
2954 local part (e.g. when creating a From: or Sender: header line), it was not
2955 quoting the local part when it contained special characters such as @.
2957 7. Installed new OpenBSD configuration files.
2959 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2960 try to make them clearer.
2962 9. Callout options, other than the timeout value, were being ignored when
2963 verifying sender addresses in header lines. For example, when using
2965 verify = header_sender/callout=no_cache
2967 the cache was (incorrectly) being used.
2969 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2970 only the Cygwin environment.
2972 11. When return_path_on_delivery was set as a log selector, if different remote
2973 addresses in the same message used different return paths and parallel
2974 remote delivery occurred, the wrong values would sometimes be logged.
2975 (Whenever a remote delivery process finished, the return path value from
2976 the most recently started remote delivery process was logged.)
2978 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2979 lines when AUTH and/or TLS are in use. This is the "received protocol"
2980 field. Exim used to use "asmtp" for authenticated SMTP, without any
2981 indication (in the protocol name) for TLS use. Now it follows the RFC and
2982 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2983 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2984 names appear in log lines as well as in Received: header lines.
2986 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2987 extensions, and comparison tests, and to fix some bugs.
2989 14. Changes to the "personal" filter test:
2991 (1) The test was buggy in that it was just doing the equivalent of
2992 "contains" tests on header lines. For example, if a user's address was
2993 anne@some.where, the "personal" test would incorrectly be true for
2995 To: susanne@some.where
2997 This test is now done by extracting each address from the header in turn,
2998 and checking the entire address. Other tests that are part of "personal"
2999 are now done using regular expressions (for example, to check local parts
3000 of addresses in From: header lines).
3002 (2) The list of non-personal local parts in From: addresses has been
3003 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3004 taken from the Sieve specification recommendations.
3006 (3) If the message contains any header line starting with "List-" it is
3007 treated as non-personal.
3009 (4) The test for "circular" in the Subject: header line has been removed
3010 because it now seems ill-conceived.
3012 15. Minor typos in src/EDITME comments corrected.
3014 16. Installed latest exipick from John Jetmore.
3016 17. If headers_add on a router specified a text string that was too long for
3017 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3018 of string_sprintf() is now avoided.
3020 18. $message_body_size was not set (it was always zero) when running the DATA
3021 ACL and the local_scan() function.
3023 19. For the "mail" command in an Exim filter, no default was being set for
3024 the once_repeat time, causing a random time value to be used if "once" was
3025 specified. (If the value happened to be <= 0, no repeat happened.) The
3026 default is now 0s, meaning "never repeat". The "vacation" command was OK
3027 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3028 (I found it when inspecting the code).
3030 20. There is now an overall timeout for performing a callout verification. It
3031 defaults to 4 times the callout timeout, which applies to individual SMTP
3032 commands during the callout. The overall timeout applies when there is more
3033 than one host that can be tried. The timeout is checked before trying the
3034 next host. This prevents very long delays if there are a large number of
3035 hosts and all are timing out (e.g. when the network connections are timing
3036 out). The value of the overall timeout can be changed by specifying an
3037 additional sub-option for "callout", called "maxwait". For example:
3039 verify = sender/callout=5s,maxwait=20s
3041 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3042 to the end before writing, but this should make it even safer).
3044 22. Exim was forgetting that it had advertised PIPELINING for the second and
3045 subsequent messages on an SMTP connection. It was also not resetting its
3046 memory on STARTTLS and an internal HELO.
3048 23. When Exim logs an SMTP synchronization error within a session, it now
3049 records whether PIPELINING has been advertised or not.
3051 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3052 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3053 rather than long int.
3055 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3057 26. Added the never_mail option to autoreply.
3063 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3064 crash if the getsockname() call failed; this can happen if a connection is
3065 closed very soon after it is established. The problem was simply in the
3066 order in which certain operations were done, causing Exim to try to write
3067 to the SMTP stream before it had set up the file descriptor. The bug has
3068 been fixed by making things happen in the correct order.
3074 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3075 before the connection was closed, thus losing the rejection response.
3077 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3078 some early Solaris releases, but causes trouble in current releases where
3079 socklen_t is defined.
3081 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3084 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3087 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3088 that releasing the top of it at the end releases what was used for sub-
3089 expansions (unless the block got too big). However, discard this block if
3090 the first thing is a variable or header, so that we can use its block when
3091 it is dynamic (useful for very large $message_headers, for example).
3093 6. Lookups now cache *every* query, not just the most recent. A new, separate
3094 store pool is used for this. It can be recovered when all lookup caches are
3095 flushed. Lookups now release memory at the end of their result strings.
3096 This has involved some general refactoring of the lookup sources.
3098 7. Some code has been added to the store_xxx() functions to reduce the amount
3099 of flapping under certain conditions.
3101 8. log_incoming_interface used to affect only the <= reception log lines. Now
3102 it causes the local interface and port to be added to several more SMTP log
3103 lines, for example "SMTP connection from", and rejection lines.
3105 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3107 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3109 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3110 could be overwritten at the end of the current message (or the start of a
3111 new message if it was set in a HELO ACL). The value is now preserved for
3112 the duration of the SMTP connection.
3114 12. If a transport had a headers_rewrite setting, and a matching header line
3115 contained an unqualified address, that address was qualified, even if it
3116 did not match any rewriting rules. The underlying bug was that the values
3117 of the flags that permit the existence of unqualified sender and recipient
3118 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3119 non-local messages, and by -bnq for local messages) were not being
3120 preserved with the message after it was received.
3122 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3123 "next input=" as part of the text comprising the host identity instead of
3124 the correct text. The code was using the same buffer for two different
3125 strings. However, depending on which order the printing function evaluated
3126 its arguments, the bug did not always show up. Under Linux, for example, my
3127 test suite worked just fine.
3129 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3130 doesn't work with some older versions of Perl. It has been changed to "my",
3131 which in any case is probably the better facility to use.
3133 15. A really picky compiler found some instances of statements for creating
3134 error messages that either had too many or two few arguments for the format
3137 16. The size of the buffer for calls to the DNS resolver has been increased
3138 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3139 for addresses that have a lot of PTR records. This alleviates a problem; it
3140 does not fully solve it.
3142 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3143 buffer now truncates the list and logs the incident, which is the same
3144 action as happens when Exim is looking up a host name and its aliases.
3145 Previously in this situation something unpredictable would happen;
3146 sometimes it was "internal error: store_reset failed".
3148 18. If a server dropped the connection unexpectedly when an Exim client was
3149 using GnuTLS and trying to read a response, the client delivery process
3150 crashed while trying to generate an error log message.
3152 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3153 string, for example:
3155 warn message = H1: something\nH2: something
3157 the text was added as a single header line from Exim's point of view
3158 though it ended up OK in the delivered message. However, searching for the
3159 second and subsequent header lines using $h_h2: did not work. This has been
3160 fixed. Similarly, if a system filter added multiple headers in this way,
3161 the routers could not see them.
3163 20. Expanded the error message when iplsearch is called with an invalid key to
3164 suggest using net-iplsearch in a host list.
3166 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3167 ACLs are no longer actually imposed (and a message to that effect is
3170 22. If a "gecos" field in a passwd entry contained escaped characters, in
3171 particular, if it contained a \" sequence, Exim got it wrong when building
3172 a From: or a Sender: header from that name. A second bug also caused
3173 incorrect handling when an unquoted " was present following a character
3174 that needed quoting.
3176 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3177 was not being matched caselessly.
3179 24. Arranged for all hyphens in the exim.8 source to be escaped with
3182 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3183 properly. Recipient callout cache records were still being keyed to include
3184 the sender, even when use_sender was set false. This led to far more
3185 callouts that were necessary. The sender is no longer included in the key
3186 when use_sender is false.
3188 26. Added "control = submission" modifier to ACLs.
3190 27. Added the ${base62d: operator to decode base 62 numbers.
3192 28. dnsdb lookups can now access SRV records.
3194 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3195 the configuration file.
3197 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3198 (-v) mode. This makes the output for a verbose queue run more intelligible.
3200 31. Added a use_postmaster feature to recipient callouts.
3202 32. Added the $body_zerocount variable, containing the number of binary zero
3203 bytes in the message body.
3205 33. The time of last modification of the "new" subdirectory is now used as the
3206 "mailbox time last read" when there is a quota error for a maildir
3209 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3211 35. Added +ignore_unknown as a special item in host lists.
3213 36. Code for decoding IPv6 addresses in host lists is now included, even if
3214 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3215 address was recognized as an IP address, but was then not correctly decoded
3216 into binary, causing unexpected and incorrect effects when compared with
3223 1. Very minor rewording of debugging text in manualroute to say "list of
3224 hosts" instead of "hostlist".
3226 2. If verify=header_syntax was set, and a header line with an unqualified
3227 address (no domain) and a large number of spaces between the end of the
3228 name and the colon was received, the reception process suffered a buffer
3229 overflow, and (when I tested it) crashed. This was caused by some obsolete
3230 code that should have been removed. The fix is to remove it!
3232 3. When running in the test harness, delay a bit after writing a bounce
3233 message to get a bit more predictability in the log output.
3235 4. Added a call to search_tidyup() just before forking a reception process. In
3236 theory, someone could use a lookup in the expansion of smtp_accept_max_
3237 per_host which, without the tidyup, could leave open a database connection.
3239 5. Added the variables $recipient_data and $sender_data which get set from a
3240 lookup success in an ACL "recipients" or "senders" condition, or a router
3241 "senders" option, similar to $domain_data and $local_part_data.
3243 6. Moved the writing of debug_print from before to after the "senders" test
3246 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3247 problems for message scanning, either using a data ACL, or using
3248 local_scan() because the Received: header was not generated till after they
3249 were called (in order to set the time as the time of reception completion).
3250 I have revised the way this works. The header is now generated after the
3251 body is received, but before the ACL or local_scan() are called. After they
3252 are run, the timestamp in the header is updated.
3258 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3259 before starting a queue runner without re-exec. This happened only when
3260 deliver_drop_privilege was set or when the Exim user was set to root. The
3261 effect of the bug was that timeouts during subsequent deliveries caused
3262 crashes instead of being properly handled. The handler is now left at its
3263 default (and expected) setting.
3265 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3266 message, again when deliver_drop_privilege is set or Exim is run as root.
3267 The bug described in (1) was not present in this case, but the tidying up
3268 of the other signals was missing. I have made the two cases consistent.
3270 3. The ignore_target_hosts setting on a manualroute router was being ignored
3271 for hosts that were looked up using the /MX notation.
3273 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3276 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3277 operated on the sender address. After changing the $sender_address to <>
3278 for the sender address verify, Exim was re-instated it as the original
3279 (before rewriting) address, but remembering that it had rewritten it, so it
3280 wasn't rewriting it again. This bug also had the effect of breaking the
3281 sender address verification caching when the sender address was rewritten.
3283 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3284 This has been changed so that if the ip literal address matches
3285 ignore_target_hosts, the router declines.
3287 7. Added expansion conditions match_domain, match_address, and match_local_
3288 part (NOT match_host).
3290 8. The placeholder for the Received: header didn't have a length field set.
3292 9. Added code to Exim itself and to exim_lock to test for a specific race
3293 condition that could lead to file corruption when using MBX delivery. The
3294 issue is with the lockfile that is created in /tmp. If this file is removed
3295 after a process has opened it but before that process has acquired a lock,
3296 there is the potential for a second process to recreate the file and also
3297 acquire a lock. This could lead to two Exim processes writing to the file
3298 at the same time. The added code performs the same test as UW imapd; it
3299 checks after acquiring the lock that its file descriptor still refers to
3300 the same named file.
3302 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3303 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3306 11. Added the smtp_active_hostname option. If used, this will typically be made
3307 to depend on the incoming interface address. Because $interface_address is
3308 not set up until the daemon has forked a reception process, error responses
3309 that can happen earlier (such as "too many connections") no longer contain
3312 12. If an expansion in a condition on a "warn" statement fails because a lookup
3313 defers, the "warn" statement is abandoned, and the next ACL statement is
3314 processed. Previously this caused the whole ACL to be aborted.
3316 13. Added the iplsearch lookup type.
3318 14. Added ident_timeout as a log selector.
3320 15. Added tls_certificate_verified as a log selector.
3322 16. Added a global option tls_require_ciphers (compare the smtp transport
3323 option of the same name). This controls incoming TLS connections.
3325 17. I finally figured out how to make tls_require_ciphers do a similar thing
3326 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3327 before starting the TLS session.
3329 18. Tabs are now shown as \t in -bP output.
3331 19. If the log selector return_path_on_delivery was set, Exim crashed when
3332 bouncing a message because it had too many Received: header lines.
3334 20. If two routers both had headers_remove settings, and the first one included
3335 a superfluous trailing colon, the final name in the first list and the
3336 first name in the second list were incorrectly joined into one item (with a
3337 colon in the middle).
3343 1. Added -C and -D options to the exinext utility, mainly to make it easier
3344 to include in the automated testing, but these could be helpful when
3345 multiple configurations are in use.
3347 2. The exinext utility was not formatting the output nicely when there was
3348 an alternate port involved in the retry record key, nor when there was a
3349 message id as well (for retries that were specific to a specific message
3350 and a specific host). It was also confused by IPv6 addresses, because of
3351 the additional colons they contain. I have fixed the IPv4 problem, and
3352 patched it up to do a reasonable job for IPv6.
3354 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3355 delivery, the log line now contains "pipelined" if PIPELINING was used.
3357 4. An SMTP transport process used to panic and die if the bind() call to set
3358 an explicit outgoing interface failed. This has been changed; it is now
3359 treated in the same way as a connect() failure.
3361 5. A reference to $sender_host_name in the part of a conditional expansion
3362 that was being skipped was still causing a DNS lookup. This no longer
3365 6. The def: expansion condition was not recognizing references to header lines
3366 that used bh_ and bheader_.
3368 7. Added the _cache feature to named lists.
3370 8. The code for checking quota_filecount in the appendfile transport was
3371 allowing one more file than it should have been.
3373 9. For compatibility with Sendmail, the command line option
3381 and sets the incoming protocol and host name (for trusted callers). The
3382 host name and its colon can be omitted when only the protocol is to be set.
3383 Note the Exim already has two private options, -pd and -ps, that refer to
3384 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3385 "s", but I don't think that's a major issue.
3387 10. A number of refactoring changes to the code, none of which should affect
3390 (a) The number of logging options was getting close to filling up the
3391 32-bit word that was used as a bit map. I have split them into two classes:
3392 those that are passed in the argument to log_write(), and those that are
3393 only ever tested independently outside of that function. These are now in
3394 separate 32-bit words, so there is plenty of room for expansion again.
3395 There is no change in the user interface or the logging behaviour.
3397 (b) When building, for example, log lines, the code previously used a
3398 macro that called string_cat() twice, in order to add two strings. This is
3399 not really sufficiently general. Furthermore, there was one instance where
3400 it was actually wrong because one of the argument was used twice, and in
3401 one call a function was used. (As it happened, calling the function twice
3402 did not affect the overall behaviour.) The macro has been replaced by a
3403 function that can join an arbitrary number of extra strings onto a growing
3406 (c) The code for expansion conditions now uses a table and a binary chop
3407 instead of a serial search (which was left over from when there were very
3408 few conditions). Also, it now recognizes conditions like "pam" even when
3409 the relevant support is not compiled in: a suitably worded error message is
3410 given if an attempt is made to use such a condition.
3412 11. Added ${time_interval:xxxxx}.
3414 12. A bug was causing one of the ddress fields not to be passed back correctly
3415 from remote delivery subprocesses. The field in question was not being
3416 subsequently used, so this caused to problems in practice.
3418 13. Added new log selectors queue_time and deliver_time.
3420 14. Might have fixed a bug in maildirsizefile handling that threw up
3421 "unexpected character" debug warnings, and recalculated the data
3422 unnecessarily. In any case, I expanded the warning message to give more
3425 15. Added the message "Restricted characters in address" to the statements in
3426 the default ACL that block characters like @ and % in local parts.
3428 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3429 Three changes have been made:
3431 (a) There was a serious bug; a negative response to MAIL caused the whole
3432 recipient domain to be cached as invalid, thereby blocking all messages
3433 to all local parts at the same domain, from all senders. This bug has
3434 been fixed. The domain is no longer cached after a negative response to
3435 MAIL if the sender used is not empty.
3437 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3440 (c) A new callout option, "use_sender" has been added for people who want
3441 the modified behaviour.
3447 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3450 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3451 indeed breaks things for older releases.
3453 3. Added additional logging to the case where there is a problem reading data
3454 from a filter that is running in a subprocess using a pipe, in order to
3455 try to track down a specific problem.
3457 4. Testing facility fudge: when running in the test harness and attempting
3458 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3459 getting "No route to host". Convert this to a timeout.
3461 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3464 6. Some OS don't have socklen_t but use size_t instead. This affects the
3465 fifth argument of getsockopt() amongst other things. This is now
3466 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3467 can be set for individual OS. I have set it for SunOS5, OSF1, and
3468 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3469 some earlier ones do not.
3471 7. Change 4.30/15 was not doing the test caselessly.
3473 8. The standard form for an IPv6 address literal was being rejected by address
3474 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3475 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3476 this, as well as the form without the "IPv6" on the front (but only when
3477 address literals are enabled, of course).
3479 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3481 10. Exim crashed if a message with an empty sender address specified by -f
3482 encountered a router with an errors_to setting. This could be provoked only
3483 by a command such as
3487 where an empty string was supplied; "<>" did not hit this bug.
3489 11. Installed PCRE release 4.5.
3491 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3492 remained set. It is now erased.
3494 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3495 times from message ids (which are base 36 rather than the normal 62).
3497 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3498 were being counted as actual protocol errors, and logged if the log
3499 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3500 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3501 and DATA following a set of rejected RCPTs do not count as protocol errors.
3502 In other words, Exim assumes they were pipelined, though this may not
3503 actually be the case. Of course, in all cases the client gets an
3504 appropriate error code.
3506 15. If a lookup fails in an ACL condition, a message about the failure may
3507 be available; it is used if testing the ACL cannot continue, because most
3508 such messages specify what the cause of the deferral is. However, some
3509 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3510 that caused an old message to be retained and used if a later statement
3511 caused a defer, replacing the real cause of the deferral.
3513 16. If an IP address had so many PTR records that the DNS lookup buffer
3514 was not large enough to hold them, Exim could crash while trying to process
3515 the truncated data. It now detects and logs this case.
3517 17. Further to 4.21/58, another change has been made: if (and only if) the
3518 first line of a message (the first header line) ends with CRLF, a bare LF
3519 in a subsequent header line has a space inserted after it, so as not to
3520 terminate the header.
3522 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3523 unnecessarily, used atoi() instead of strtol(), and didn't check the
3524 termination when getting file sizes from file names by regex.
3526 19. Completely re-implemented the support for maildirsize files, in the light
3527 of a number of problems with the previous contributed implementation
3528 (4.30/29). In particular:
3530 . If the quota is zero, the maildirsize file is maintained, but no quota is
3533 . If the maildir directory does not exist, it is created before any attempt
3534 to write a maildirsize file.
3536 . The quota value in the file is just a cache; if the quota is changed in
3537 the transport, the new value overrides.
3539 . A regular expression is available for excluding directories from the
3542 20. The autoreply transport checks the characters in options that define the
3543 message's headers; it allows continued headers, but it was checking with
3544 isspace() after an embedded newline instead of explicitly looking for a
3547 21. If all the "regular" hosts to which an address was routed had passed their
3548 expiry times, and had not reached their retry times, the address was
3549 bounced, even if fallback hosts were defined. Now Exim should go on to try
3552 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3553 equivalent code in the SMTP transport. Some hosts send humungous responses
3554 to HELO/EHLO, more than 1024 it seems.
3556 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3557 gives compiler warnings in some environments. I've now done it "properly",
3560 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3561 (because of problems with the built-in one) was declared to return uschar *
3562 instead of char *, causing compiler failure.
3564 25. Fixed a file descriptor leak when processing alias/forward files.
3566 26. Fixed a minor format string issue in dbfn.c.
3568 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3570 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3571 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3574 29. When a sender address is verified, it is cached, to save repeating the test
3575 when there is more than one recipient in a message. However, when the
3576 verification involves a callout, it is possible for different callout
3577 options to be set for different recipients. It is too complicated to keep
3578 track of this in the cache, so now Exim always runs a verification when a
3579 callout is required, relying on the callout cache for the optimization.
3580 The overhead is duplication of the address routing, but this should not be
3583 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3584 to be verified with callout=postmaster, and the main callout worked but the
3585 postmaster check failed, the verification correctly failed. However, if a
3586 subsequent RCPT command asked for sender verification *without* the
3587 postmaster check, incorrect caching caused this verification also to fail,
3590 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3591 it was not caching the DNS options (qualify_single, search_parents) that
3592 were used when the lookup failed. A subsequent lookup with different
3593 options therefore always gave the same answer, though there were cases
3594 where it should not have. (Example: a "domains = !$mx_any" option on a
3595 dnslookup router: the "domains" option is always processed without any
3596 widening, but the router might have qualify_single set.) Now Exim uses the
3597 cached value only when the same options are set.
3599 32. Added John Jetmore's "exipick" utility to the distribution.
3601 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3602 than a timeout (e.g. a certificate is required, and is not provided), an
3603 Exim server now closes the connection immediately. Previously it waited for
3604 the client to close - but if the client is SSL, it seems that they each
3605 wait for each other, leading to a delay before one of them times out.
3607 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3608 maintained 0.8.x compatibility because I don't think many are using it, and
3609 it is clearly obsolete.
3611 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3614 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3615 client certificate was expired. A simple patch fixes this, though I don't
3616 understand the full logic of why the verify callback is called multiple
3619 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3620 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3621 which causes problems with some clients (such as the Certicom SSL Plus
3622 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3623 disables the coutermeasure allowing Eudora to connect."
3625 38. Exim was not checking that a write() to a log file succeeded. This could
3626 lead to Bad Things if a log got too big, in particular if it hit a file
3627 size limit. Exim now panics and dies if it cannot write to a log file, just
3628 as it does if it cannot open a log file.
3630 39. Modified OS/Makefile-Linux so that it now contains
3632 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3634 The two -D definitions ensure that Exim is compiled with large file
3635 support, which makes it possible to handle log files that are bigger than
3638 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3639 instance) a domain was checked against a named list that involved a lookup,
3640 causing $domain_data to be set, then another domain was checked against the
3641 same list, then the first domain was re-checked, the value of $domain_data
3642 after the final check could be wrong. In particular, if the second check
3643 failed, it could be set empty. This bug probably also applied to
3646 41. The strip_trailing_dot option was not being applied to the address given
3647 with the -f command-line option.
3649 42. The code for reading a message's header from the spool was incrementing
3650 $received_count, but never initializing it. This meant that the value was
3651 incorrect (doubled) while delivering a message in the same process in which
3652 it was received. In the most common configuration of Exim, this never
3653 happens - a fresh exec is done - but it can happen when
3654 deliver_drop_privilege is set.
3656 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3657 it now includes up to 150 characters of the unexpected data in the log
3660 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3661 and building data strings. The size of both of these buffers was 10 000
3662 bytes - far larger than anybody would *ever* want, thought I. Needless to
3663 say, somebody hit the limit. I have increased the maximum line length to
3664 20 000 and the maximum data length of concatenated lines to 100 000. I have
3665 also fixed two bugs, because there was no checking on these buffers. Tsk,
3666 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3667 buffer is too small.
3669 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3670 lsearch lookups. Now it does.
3672 46. When parsing a route_list item in a manualroute router, a fixed-length
3673 buffer was used for the list of hosts. I made this 1024 bytes long,
3674 thinking that nobody would ever have a list of hosts that long. Wrong.
3675 Somebody had a whole pile of complicated expansion conditions, and the
3676 string was silently truncated, leading to an expansion error. It turns out
3677 that it is easier to change to an unlimited length (owing to other changes
3678 that have happened since this code was originally written) than to build
3679 structure for giving a limitation error. The length of the item that
3680 expands into the list of hosts is now unlimited.
3682 47. The lsearch lookup could not handle data where the length of text line was
3683 more than 4095 characters. Such lines were truncated, leading to shortened
3684 data being returned. It should now handle lines of any length.
3686 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3687 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3690 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3691 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3692 no longer happens in comment lines. A list of replacements is now placed
3693 at the head of all of the source files, except those whose only change is
3694 to replace PERL_COMMAND in the very first #! line.
3696 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3697 messages on the queue, with a bottom-up merge sort, using code contributed
3698 by Michael Haardt. This should make operations like -bp somewhat faster on
3699 large queues. It won't affect queue runners, except when queue_run_in_order
3702 51. Installed eximstats 1.31 in the distribution.
3704 52. Added support for SRV lookups to the dnslookup router.
3706 53. If an ACL referred to $message_body or $message_body_end, the value was not
3707 reset for any messages that followed in the same SMTP session.
3709 54. The store-handling optimization for building very long strings was not
3710 differentiating between the different store pools. I don't think this
3711 actually made any difference in practice, but I've tidied it.
3713 55. While running the routers to verify a sender address, $sender_address
3714 was still set to the sender address. This is wrong, because when routing to
3715 send a bounce to the sender, it would be empty. Therefore, I have changed
3716 it so that, while verifying a sender address, $sender_address is set to <>.
3717 (There is no change to what happens when verifying a recipient address.)
3719 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3720 target A or AAAA records (if not already returned) without resetting the
3721 qualify_single or search_parents options of the DNS resolver. These are
3722 inappropriate in this case because the targets of MX and SRV records must
3723 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3724 target that, when qualified, matched something in the local domain. These
3725 two options are now turned off when doing these lookups.
3727 57. It seems that at least some releases of Reiserfs (which does not have the
3728 concept of a fixed number of inodes) returns zero and not -1 for the
3729 number of available inodes. This interacted badly with check_spool_inodes,
3730 which assumed that -1 was the "no such thing" setting. What I have done is
3731 to check that the total number of inodes is greater than zero before doing
3732 the test of how many are available.
3734 58. When a "warn" ACL statement has a log_message modifier, the message is
3735 remembered, and not repeated. This is to avoid a lot of repetition when a
3736 message has many recipients that cause the same warning to be written.
3737 Howewer, Exim was preserving the list of already written lines for an
3738 entire SMTP session, which doesn't seem right. The memory is now reset if a
3739 new message is started.
3741 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3742 debugging output unless log_rewrite was also set.
3744 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3745 of (int)(handle) when we know that handle contains (void *)(-1).
3747 61. The Exim daemon panic-logs an error return when it closes the incoming
3748 connection. However "connection reset by peer" seems to be common, and
3749 isn't really an error worthy of noting specially, so that particular error
3752 62. When Exim is trying to find all the local interfaces, it used to panic and
3753 die if the ioctl to get the interface flags failed. However, it seems that
3754 on at least one OS (Solaris 9) it is possible to have an interface that is
3755 included in the list of interfaces, but for which you get a failure error
3756 for this call. This happens when the interface is not "plumbed" into a
3757 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3758 failure of the "get flags" call assumes that the interface is down.
3760 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3761 makes life easier for people who are doing arithmetic on fields extracted
3762 from dates, where you often get leading zeros that should not be
3763 interpreted as octal.
3765 64. Added qualify_domain to the redirect router, to override the global
3768 65. If a pathologically long header line contained very many addresses (the
3769 report of this problem mentioned 10 000) and each of them was rewritten,
3770 Exim could use up a very large amount of memory. (It kept on making new
3771 copies of the header line as it rewrote, and never released the old ones.)
3772 At the expense of a bit more processing, the header rewriting function has
3773 been changed so that it no longer eats memory in this way.
3775 66. The generation of the Received: header has been moved from the time that a
3776 message starts to be received, to the time that it finishes. The timestamp
3777 in the Received: header should now be very close to that of the <= log
3778 line. There are two side-effects of this change:
3780 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3781 logged header lines no longer include the local Received: line, because
3782 it has not yet been created. The same applies to a copy of the message
3783 that is returned to a non-SMTP sender when a message is rejected.
3785 (b) When a filter file is tested using -bf, no additional Received: header
3786 is added to the test message. After some thought, I decided that this
3789 This change does not affect the value of $received_for. It is still set
3790 after address rewriting, but before local_scan() is called.
3792 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3794 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3795 gave an unhelpful panic error message, and a defer error. I have managed to
3796 change this behaviour so that it now rejects any supplied certificate,
3797 which seems right, as the list of acceptable certificates is empty.
3799 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3800 gave an unhelpful defer error. I have not managed to make this reject any
3801 supplied certificates, but the error message it gives is "no certificate
3802 supplied", which is not helpful.
3804 70. exigrep's output now also includes lines that are not associated with any
3805 message, but which match the given pattern. Implemented by a patch from
3806 Martin Sluka, which also tidied up the Perl a bit.
3808 71. Recipient callout verification, like sender verification, was using <> in
3809 the MAIL FROM command. This isn't really the right thing, since the actual
3810 sender may affect whether the remote host accepts the recipient or not. I
3811 have changed it to use the actual sender in the callout; this means that
3812 the cache record is now keyed on a recipient/sender pair, not just the
3813 recipient address. There doesn't seem to be a real danger of callout loops,
3814 since a callout by the remote host to check the sender would use <>.
3815 [SEE ABOVE: changed after hitting problems.]
3817 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3818 temporary errors. However, in the case of such a code being given after
3819 the end of a data transmission (i.e. after ".") Exim was failing to write
3820 a retry record for the message. (Yes, there was some broken host that was
3821 actually sending 8xx at this point.)
3823 73. An unknown lookup type in a host list could cause Exim to panic-die when
3824 the list was checked. (An example that provoked this was putting <; in the
3825 middle of a list instead of at the start.) If this happened during a DATA
3826 ACL check, a -D file could be left lying around. This kind of configuration
3827 error no longer causes Exim to die; instead it causes a defer errror. The
3828 incident is still logged to the main and panic logs.
3830 74. Buglet left over from Exim 3 conversion. The message "too many messages
3831 in one connection" was written to the rejectlog but not the mainlog, except
3832 when address rewriting (yes!) was being logged.
3834 75. Added write_rejectlog option.
3836 76. When a system filter was run not as root (that is, when system_filter_user
3837 was set), the values of the $n variables were not being returned to the
3838 main process; thus, they were not subsequently available in the $sn
3841 77. Added +return_path_on_delivery log selector.
3843 78. A connection timeout was being treated differently from recipients deferred
3844 when testing hosts_max_try with a message that was older than the host's
3845 retry timeout. (The host should not be counted, thus allowing all hosts to
3846 be tried at least once before bouncing.) This may have been the cause of an
3847 occasionally reported bug whereby a message would remain on the queue
3848 longer than the retry timeout, but would be bounced if a delivery was
3849 forced. I say "may" because I never totally pinned down the problem;
3850 setting up timeout/retry tests is difficult. See also the next item.
3852 79. The ultimate address timeout was not being applied to errors that involved
3853 a combination of host plus message (for example, a timeout on a MAIL
3854 command). When an address resolved to a number of possible hosts, and they
3855 were not all tried for each delivery (e.g. because of hosts_max_try), a
3856 message could remain on the queue longer than the retry timeout.
3858 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3861 81. Fixed an obscure SMTP outgoing bug which required at least the following
3862 conditions: (a) there was another message waiting for the same server;
3863 (b) the server returned 5xx to all RCPT commands in the first message so
3864 that the message was not completed; (c) the server dropped the connection
3865 or gave a negative response to the RSET that Exim sends to abort the
3866 transaction. The observed case was a dropped connection after DATA that had
3867 been sent in pipelining mode. That is, the server had advertised PIPELINING
3868 but was not implementing it correctly. The effect of the bug was incorrect
3869 behaviour, such as trying another host, and this could lead to a crash.
3875 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3876 and daemon.c were passed as pointers to ints; they should have been
3877 pointers to socklen_t variables (which are typically unsigned ints).
3879 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3882 3. Fixed a really odd bug that affected only the testing scheme; patching a
3883 certain fixed string in the binary changed the value of another string that
3884 happened to be identical to the end of the original first string.
3886 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3887 name", it returns that address as the IP address. On some operating
3888 systems (e.g. Solaris), it also passes back the IP address string as the
3889 "host name". However, on others (e.g. Linux), it passes back an empty
3890 string. Exim wasn't checking for this, and was changing the host name to an
3891 empty string, assuming it had been canonicized.
3893 5. Although rare, it is permitted to have more than one PTR record for a given
3894 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3895 all the names associated with an address, because they do in Solaris.
3896 However, it seems that they do not in Linux for data that comes from the
3897 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3898 I found this out when I moved to a new Linux workstation and tried to run
3899 the Exim test suite.
3901 To get round this problem I have changed the code so that it now does its
3902 own call to the DNS to look up PTR records when searching for a host name.
3903 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3904 addresses that are only in /etc/hosts are still found.
3906 This behaviour is, however, controlled by an option called host_lookup_
3907 order, which defaults to "bydns:byaddr". If people want to use the other
3908 order, or indeed, just use one or the other means of lookup, they can
3909 specify it in this variable.
3911 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3912 some operating systems, this comes back from gethostbyaddr() as an empty
3913 string, and this is what Exim used to test for. However, it seems that in
3914 other systems, "." is yielded. Exim now tests for this case too.
3916 7. The values of check_spool_space and check_log_space are now held internally
3917 as a number of kilobytes instead of an absolute number of bytes. If a
3918 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3919 kilobyte. This means that much larger values can be stored.
3921 8. Exim monitor: an attempt to get the action menu when not actually pointing
3922 at a message produces an empty menu entitled "No message selected". This
3923 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3924 no entries in it ("Shell widget menu has zero width and/or height"). So I
3925 have added a single, blank menu entry in this case.
3927 9. Added ${quote_local_part.
3929 10. MIME decoding is now applied to the contents of Subject: header lines when
3932 11. Now that a reference to $sender_host_address automatically causes a reverse
3933 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3934 host lookup before query-style lookups in lists that might use this
3935 variable. This has therefore been abolished, and the "net-" prefix is no
3936 longer necessary for query-style lookups.
3938 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3939 have been a typo for LFLAGS, so it has been changed.
3941 13. The install script calls Exim with "-C /dev/null" in order to find the
3942 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3943 to be output. Howeve, since Exim outputs its version number before the
3944 error, it didn't break the script. It just looked ugly. I fixed this by
3945 always allowing "-C /dev/null" if the caller is root.
3947 14. Ignore overlarge ACL variable number when reading spool file - insurance
3948 against a later release with more variables having written the file.
3950 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3951 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3952 this, as well as the form without the "IPv6" on the front.
3954 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3955 OS/Makefile-Darwin file.
3957 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3958 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3960 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3961 variable. However, some OS use ints for inodes. I've added cast to long int
3962 to get rid of the compiler warning.
3964 19. I had forgotten to lock out "/../" in configuration file names when
3965 ALT_CONFIG_PREFIX was set.
3967 20. Routers used for verification do not need to specify transports. However,
3968 if such a router generated a host list, and callout was configured, Exim
3969 crashed, because it could not find a port number from the (non-existent)
3970 transport. It now assumes port 25 in this circumstance.
3972 21. Added the -t option to exigrep.
3974 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3975 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3976 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3977 them is set, LOOKUP_LSEARCH is forced.
3979 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3981 24. Added sender and host information to the "rejected by local_scan()" log
3982 line; previously there was no indication of these.
3984 25. Added .include_if_exists.
3986 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3987 Linux. It turns out that not all file systems support this. Apparently some
3988 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3989 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3990 not supported on the file descriptor, is now ignored when Exim is trying to
3991 sync a directory. This applies only to Linux.
3993 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3995 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3996 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3999 29. Added support for maildirsize files from supplied patch (modified a bit).
4001 30. The use of :fail: followed by an empty string could lead Exim to respond to
4002 sender verification failures with (e.g.):
4004 550 Verification failed for <xxx>
4005 550 Sender verify failed
4007 where the first response line was missing the '-' that indicates it is not
4008 the final line of the response.
4010 31. The loop for finding the name of the user that called Exim had a hardwired
4011 limit of 10; it now uses the value of finduser_retries, which is used for
4012 all other user lookups.
4014 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4017 33. Exim was neglecting to zero errno before one call of strtol() when
4018 expanding a string and expecting an integer value. On some systems this
4019 resulted in spurious "integer overflow" errors. Also, it was casting the
4020 result into an int without checking.
4022 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4023 did not work. The code looks as if it has *never* worked, though it appears
4024 to have been documented since at least releast 1.62. I have made it work.
4026 35. The "timeout_DNS" error in retry rules, also documented since at least
4027 1.62, also never worked. As it isn't clear exactly what this means, and
4028 clearly it isn't a major issue, I have abolished the feature by treating it
4029 as "timeout", and writing a warning to the main and panic logs.
4031 36. The display of retry rules for -brt wasn't always showing the error code
4034 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4035 timeout_connect_A, timeout_connect_MX.
4037 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4038 to the empty sender.
4040 39. The daemon was not analysing the content of -oX till after it had closed
4041 stderr and disconnected from the controlling terminal. This meant that any
4042 syntax errors were only noted on the panic log, and the return code from
4043 the command was 0. By re-arranging the code a little, I've made the
4044 decoding happen first, so such errors now appear on stderr, and the return
4045 code is 1. However, the actual setting up of the sockets still happens in
4046 the disconnected process, so errors there are still only recorded on the
4049 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4050 connections (as happens on some IP stacks) was logged at start up time as
4051 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4052 it from "IPv6 and IPv4", which means that two separate sockets are being
4055 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4056 says whether AF_INET or AF_INET6 was passed as an argument.
4058 42. Exiwhat output was messed up when time zones were included in log
4061 43. Exiwhat now gives more information about the daemon's listening ports,
4062 and whether -tls-on-connect was used.
4064 44. The "port" option of the smtp transport is now expanded.
4066 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4067 silently ignored. Now an error message is written to the main and panic
4070 46. There's a new ACL modifier called "logwrite" which writes to a log file
4071 as soon as it is encountered.
4073 47. Added $local_user_uid and $local_user_gid at routing time.
4075 48. Exim crashed when trying to verify a sender address that was being
4078 49. Exim was recognizing only a space character after ".include". It now also
4079 recognizes a tab character.
4081 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4082 extracting the relevant information from the specification. The man page no
4083 longer contains scrambled data for the -d option, and I've added a section
4084 at the front about calling Exim under different names.
4086 51. Added "extra_headers" argument to the "mail" command in filter files.
4088 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4091 53. Installed eximstats 1.29.
4093 54. Added transport_filter_timeout as a generic transport option.
4095 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4096 Cc: header lines. This was required by RFC 822, but it not required by RFC
4099 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4100 incoming messages that did not have them. Now it does so only if the
4101 message originates locally, that is, if there is no associated remote host
4102 address. When Resent- header lines are present, this applies to the Resent-
4103 lines rather than the non-Resent- lines.
4105 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4106 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4108 58. Messages for configuration errors now include the name of the main
4109 configuration file - useful now that there may be more than one file in a
4110 list (.included file names were always shown).
4112 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4113 for those rare installations that do not start the daemon as root or run it
4114 setuid root. I've cut out the call to initgroups() if the daemon is not
4117 60. The Exim user and group can now be bound into the binary as text strings
4118 that are looked up at the start of Exim's processing.
4120 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4122 62. Added $mailstore_basename variable.
4124 63. Installed patch to sieve.c from Michael Haardt.
4126 64. When Exim failed to open the panic log after failing to open the main log,
4127 the original message it was trying to log was written to stderr and debug
4128 output, but if they were not available (the usual case in production), it
4129 was lost. Now it is written to syslog before the two lines that record the
4130 failures to open the logs.
4132 65. Users' Exim filters run in subprocesses under the user's uid. It is
4133 possible for a "deliver" command or an alias in a "personal" command to
4134 provoke an address rewrite. If logging of address rewriting is configured,
4135 this fails because the process is not running as root or exim. There may be
4136 a better way of dealing with this, but for the moment (because 4.30 needs
4137 to be released), I have disabled address rewrite logging when running a
4138 filter in a non-root, non-exim process.
4144 1. The buildconfig auxiliary program wasn't quoting the value set for
4145 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4146 not defined. This bug was masked in 4.22 by the effect that was fixed in
4149 2. Some messages that were rejected after a message id was allocated were
4150 shown as "incomplete" by exigrep. It no longer does this for messages that
4151 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4153 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4154 have allow_domain_literals set, the ID did not get logged in the <= line.
4155 Domain literals are now always recognized in Message-ID: header lines.
4157 4. The first argument for a ${extract expansion item is the key name or field
4158 number. Leading and trailing spaces in this item were not being ignored,
4159 causing some misleading effects.
4161 5. When deliver_drop_privilege was set, single queue runner processes started
4162 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4163 same command in the process it spins off) were not dropping privilege.
4165 6. When the daemon running as "exim" started a queue runner, it always
4166 re-executed Exim in the spun-off process. This is a waste of effort when
4167 deliver_drop_privilege is set. The new process now just calls the
4168 queue-runner function directly.
4174 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4177 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4178 ignored. Though the use of -oP was forcing the writing of a pid file, it
4179 was always written to the default place.
4181 3. If the message "no IP address found for host xxxx" is generated during
4182 incoming verification, it is now followed by identification of the incoming
4183 connection (so you can more easily find what provoked it).
4185 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4187 5. Added some features to "harden" Exim a bit more against certain attacks:
4189 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4190 be put in Local/Makefile. This is like the never_users runtime option,
4191 but it cannot be overridden. The default setting is "root".
4193 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4194 prefix string with which any file named in a -C command line option
4197 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4198 is retained for -C and -D only if the caller of Exim is root. Without
4199 it, the exim user may also use -C and -D and retain privilege.
4201 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4202 command line option is disabled.
4204 6. Macro names set by the -D option must start with an upper case letter, just
4205 like macro names defined in the configuration file.
4207 7. Added "dereference=" facility to LDAP.
4209 8. Two instances of the typo "uknown" in the source files are fixed.
4211 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4212 the Configure-Makefile script screwed up while processing it.
4214 10. Incorporated PCRE 4.4.
4216 11. The SMTP synchronization check was not operating right at the start of an
4217 SMTP session. For example, it could not catch a HELO sent before the client
4218 waited for the greeting. There is now a check for outstanding input at the
4219 point when the greeting is written. Because of the duplex, asynchronous
4220 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4221 way, but not yet received, when the check is performed.
4223 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4224 on TCP/IP sockets, because this apparently causes some broken clients to
4227 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4228 unchanged) from the Cygwin maintainer.
4230 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4231 was supported instead of testing for mbx. Effectively a typo.
4233 15. The spa authenticator server code was not checking that the input it
4234 received was valid base64.
4236 16. The debug output line for the "set" modifier in ACLs was not showing the
4237 name of the variable that was being set.
4239 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4241 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4242 lookup on its own. Something else was needed to trigger the lookup. For
4243 example, a match in host_lookup or the need for a host name in a host list.
4244 Now, if $sender_host_name is referenced and the host name has not yet been
4245 looked up, a lookup is performed. If the lookup fails, the variable remains
4246 empty, and $host_lookup_failed is set to "1".
4248 19. Added "eqi" as a case-independent comparison operator.
4250 20. The saslauthd authentication condition could segfault if neither service
4251 nor realm was specified.
4253 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4254 error message that was logged was misleading, and incoming SMTP
4255 connections were dropped. The message is now more accurate, and temporary
4256 errors are given to SMTP connections.
4258 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4259 (except RSET) with a 421 error, until QUIT is received. However, it was
4260 failing to send a response to QUIT.
4262 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4263 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4264 ACL is now run after the helo_try_verify_hosts code.
4266 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4267 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4268 case-independent, so other case variants are also recognized). Apparently
4269 some systems use these upper case variants.
4271 25. If more than two messages were waiting for the same host, and a transport
4272 filter was specified for the transport, Exim sent two messages over the
4273 same TCP/IP connection, and then failed with "socket operation on non-
4274 socket" when it tried to send the third.
4276 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4278 27. The extern definition of crypt16() in expand.c was not being excluded when
4279 the OS had its own crypt16() function.
4281 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4282 as a preferred synonym for return_size_limit, both as an option and as an
4285 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4287 30. Changed the default configuration ACL to relax the local part checking rule
4288 for addresses that are not in any local domains. For these addresses,
4289 slashes and pipe symbols are allowed within local parts, but the sequence
4290 /../ is explicitly forbidden.
4292 31. SPA server authentication was not clearing the challenge buffer before
4295 32. log_message in a "warn" ACL statement was writing to the reject log as
4296 well as to the main log, which contradicts the documentation and doesn't
4297 seem right (because no rejection is happening). So I have stopped it.
4299 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4300 However, I am unable to do any testing of this.
4302 34. Fixed an infelicity in the appendfile transport. When checking directories
4303 for a mailbox, to see if any needed to be created, it was accidentally
4304 using path names with one or more superfluous leading slashes; tracing
4305 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4307 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4308 contents are added to the log line that is written for every discarded
4309 recipient. (Previously a log_message setting was ignored.)
4311 36. The ${quote: operator now quotes the string if it is empty.
4313 37. The install script runs exim in order to find its version number. If for
4314 some reason other than non-existence or emptiness, which it checks, it
4315 could not run './exim', it was installing it with an empty version number,
4316 i.e. as "exim-". This error state is now caught, and the installation is
4319 38. An argument was missing from the function that creates an error message
4320 when Exim fails to connect to the socket for saslauthd authentication.
4321 This could cause Exim to crash, or give a corrupted message.
4323 39. Added isip, isip4, and isip6 to ${if conditions.
4325 40. The ACL variables $acl_xx are now saved with the message, and can be
4326 accessed later in routers, transports, and filters.
4328 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4329 strings in the file are not string-expanded.
4331 42. If a MAIL command specified a SIZE value that was too large to fit into an
4332 int variable, the check against message_size_limit failed. Such values are
4333 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4334 day this will have to be increased, but I don't think I want to be around
4335 when emails are that large.
4342 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4343 iconv() is not standard in FreeBSD.
4345 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4346 IPv6 enabled. The observed symptom was a segmentation fault on return from
4347 the function os_common_find_running_interfaces() in src/os.c.
4349 3. In the check_special_case() function in daemon.c I had used "errno" as an
4350 argument name, which causes warnings on some systems. This was basically a
4351 typo, since it was named "eno" in the comments!
4353 4. The code that waits for the clock to tick (at a resolution of some fraction
4354 of a second) so as to ensure message-id uniqueness was always waiting for
4355 at least one whole tick, when it could have waited for less. [This is
4356 almost certainly not relevant at current processor speeds, where it is
4357 unlikely to ever wait at all. But we try to future-proof.]
4359 5. The function that sleeps for a time interval that includes fractions of a
4360 second contained a race. It did not block SIGALRM between setting the
4361 timer, and suspending (a couple of lines later). If the interval was short
4362 and the sigsuspend() was delayed until after it had expired, the suspension
4363 never ended. On busy systems this could lead to processes getting stuck for
4366 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4367 process, before it forks any delivery processes. The open lookup caching
4368 mechanism meant that the open file or database connection was passed into
4369 the delivery process. The problem was that delivery processes always tidy
4370 up cached lookup data. This could cause a problem for the next delivery
4371 process started by the queue runner, because the external queue runner
4372 process does not know about the closure. So the next delivery process
4373 still has data in the lookup cache. In the case of a file lookup, there was
4374 no problem because closing a file descriptor in a subprocess doesn't affect
4375 the parent. However, if the lookup was caching a connection to a database,
4376 the connection was closed, and the second delivery process was likely to
4377 see errors such as "PGSQL: query failed: server closed the connection
4378 unexpectedly". The problem has been fixed by closing all cached lookups
4379 in a queue runner before running a delivery process.
4381 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4382 seem to have the "const" qualifier which it has on other OS. I've
4385 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4386 *of the same type* (client or server) with the same public name that an
4387 error should be diagnosed.
4389 9. When Exim looked up a host name for an IP address, but failed to find the
4390 original IP address when looking up the host name (a safety check), it
4391 output the message "<ip address> does not match any IP for NULL", which was
4392 confusing, to say the least. The bug was that the host name should have
4393 appeared instead of "NULL".
4395 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4396 user that is built into the binary, and the -C or -D options is used, root
4397 privilege is dropped before the configuration file is read. In addition,
4398 logging is switched to stderr instead of the normal log files. If the
4399 configuration then re-defines the Exim user, the unprivileged environment
4400 is probably not what is expected, so Exim logs a panic warning message (but
4403 However, if deliver_drop_privilege is set, the unprivileged state may well
4404 be exactly what is intended, so the warning has been cut out in that case,
4405 and Exim is allowed to try to write to its normal log files.
4411 1. smtp_return_error_details was not giving details for temporary sender
4412 or receiver verification errors.
4414 2. Diagnose a configuration error if two authenticators have the same public
4417 3. Exim used not to create the message log file for a message until the first
4418 delivery attempt. This could be confusing when incoming messages were held
4419 for policy or load reasons. The message log file is now created at the time
4420 the message is received, and an initial "Received" line is written to it.
4422 4. The automatically generated man page for command line options had a minor
4423 bug that caused no ill effects; however, a more serious problem was that
4424 the procedure for building the man page automatically didn't always
4425 operate. Consequently, release 4.20 contains an out-of-date version. This
4426 shouldn't happen again.
4428 5. When building Exim with embedded Perl support, the script that builds the
4429 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4430 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4432 6. The freeze_tell option was not being used for messages that were frozen on
4433 arrival, either by an ACL or by local_scan().
4435 7. Added the smtp_incomplete_transaction log selector.
4437 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4438 was accepting AUTH without a new EHLO.
4440 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4441 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4444 10. Logging of TCP/IP connections (when configured) now happens in the main
4445 daemon process instead of the child process, so that the TCP/IP connection
4446 count is more accurate (but it can never be perfect).
4448 11. The use of "drop" in a nested ACL was not being handled correctly in the
4449 outer ACL. Now, if condition failure induced by the nested "drop" causes
4450 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4451 or "require"), the connection is dropped.
4453 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4454 that yield "discard" can now be used with an "accept" or a "discard" verb,
4455 but an error is generated for any others (because I can't see a useful way
4456 to define what should happen).
4458 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4459 are now processed in the same way as lines in the Exim configuration file.
4460 In particular, continuation lines are supported.
4462 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4464 15. Added -ti meaning -t -i.
4466 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4467 domains, and warn by logging if others are found.
4469 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4470 returned from GIFCONF when Exim is trying to find the list of interfaces on
4471 a host. The code in os.c has been modified to copy each ifreq to an aligned
4472 structure in all cases.
4474 Also, in some cases, the returned ifreq's were being copied to a 'struct
4475 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4476 means the last couple of bytes of an IPv6 address could be chopped if the
4477 ifreq contained only a normal sockaddr (14 bytes storage).
4479 18. Named domain lists were not supported in the hosts_treat_as_local option.
4480 An entry such as +xxxx was not recognized, and was treated as a literal
4483 19. Ensure that header lines added by a DATA ACL are included in the reject log
4484 if the ACL subsequently rejects the message.
4486 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4487 MD5 (which is deprecated).
4489 21. When testing a filter file using -bf, Exim was writing a message when it
4490 took the sender from a "From " line in the message, but it was not doing so
4491 when it took $return_path from a Return-Path: header line. It now does.
4493 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4494 with a valid header line field name (a series of printing characters
4495 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4497 23. Changed "disc" in the source to "disk" to conform to the documentation and
4498 the book and for uniformity.
4500 24. Ignore Sendmail's -Ooption=value command line item.
4502 25. When execve() failed while trying to run a command in a pipe transport,
4503 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4504 could be confused with a return value of 69 from the command itself. This
4505 has been changed to 127, the value the shell returns if it is asked to run
4506 a non-existent command. The wording for the related log line suggests a
4507 non-existent command as the problem.
4509 26. If received_header_text expands to an empty string, do not add a Received:
4510 header line to the message. (Well, it adds a token one on the spool, but
4511 marks it "old" so that it doesn't get used or transmitted.)
4513 27. Installed eximstats 1.28 (addition of -nt option).
4515 28. There was no check for failure on the call to getsockname() in the daemon
4516 code. This can fail if there is a shortage of resources on the system, with
4517 ENOMEM, for example. A temporary error is now given on failure.
4519 29. Contrary to the C standard, it seems that in some environments, the
4520 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4521 program. Exim now does this explicitly; it affects the formatting of
4522 timestamps using strftime().
4524 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4525 complaints. I've now initialized all the variables, to avoid this.
4527 32. Header lines added by a system filter were not being "seen" during
4528 transport-time rewrites.
4530 33. The info_callback() function passed to OpenSSL is set up with type void
4531 (*)(SSL *, int, int), as described somewhere. However, when calling the
4532 function (actually a macro) that sets it up, the type void(*)() is
4533 expected. I've put in a cast to prevent warnings from picky compilers.
4535 34. If a DNS black list lookup found a CNAME record, but there were no A
4536 records associated with the domain it pointed at, Exim crashed.
4538 35. If a DNS black list lookup returned more than one A record, Exim ignored
4539 all but the first. It now scans all returned addresses if a particular IP
4540 value is being sought. In this situation, the contents of the
4541 $dnslist_value variable are a list of all the addresses, separated by a
4544 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4545 to accept a host name and all its aliases if the forward lookup for any of
4546 them yielded the IP address of the incoming connection. Now it accepts only
4547 those names whose forward lookup yields the correct IP address. Any other
4548 names are discarded. This closes a loophole whereby a rogue DNS
4549 administrator could create reverse DNS records to break through a
4550 wildcarded host restriction in an ACL.
4552 37. If a user filter or a system filter that ran in a subprocess used any of
4553 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4554 the wrong values were passed to the pipe command ($thisaddress had the
4555 value of $0, $0 had the value of $1, etc). This bug was introduced by
4556 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4558 38. Improved the line breaking for long SMTP error messages from ACLs.
4559 Previously, if there was no break point between 40 and 75 characters, Exim
4560 left the rest of the message alone. Two changes have been made: (a) I've
4561 reduced the minimum length to 35 characters; (b) if it can't find a break
4562 point between 35 and 75 characters, it looks ahead and uses the first one
4563 that it finds. This may give the occasional overlong line, but at least the
4564 remaining text gets split now.
4566 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4567 file descriptors might be low, and that setting 1000 would always raise it.
4568 It turns out that in some environments, the limit is already over 1000 and
4569 that lowering it causes trouble. So now Exim takes care not to decrease it.
4571 40. When delivering a message, the value of $return_path is set to $sender_
4572 address at the start of routing (routers may change the value). By an
4573 oversight, this default was not being set up when an address was tested by
4574 -bt or -bv, which affected the outcome if any router or filter referred to
4577 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4578 log only when "message" or "log_message" are set. However, if one of the
4579 conditions was an address verification, or a call to a nested ACL, the
4580 messages generated by the underlying test were being passed through. This
4581 no longer happens. The underlying message is available in $acl_verify_
4582 message for both "message" and "log_message" expansions, so it can be
4583 passed through if needed.
4585 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4586 new expansion $bh_ to give the encoded byte string without charset
4587 translation. Translation happens only if iconv() is available; HAVE_ICONV
4588 indicates this at build time. HEADERS_CHARSET gives the charset to
4589 translate to; headers_charset can change it in the configuration, and
4590 "headers charset" can change it in an individual filter file.
4592 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4593 that creates RFC 2047 encoded "words" labels them as that charset instead
4594 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4595 expansion operator; (ii) when Exim creates a From: line for a local
4596 message; (iii) when a header line is rewritten to include a "phrase" part.
4598 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4599 buggy, causing it to skip the first lines of messages whose message ID
4600 ended in 'D'. This would not have bitten before Exim release 4.14, because
4601 message IDs were unlikely to end in 'D' before then. The effect was to have
4602 incorrect size information for certain domains.
4604 45. #include "config.h" was missing at the start of the crypt16.c module. This
4605 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4608 46. If there was a timeout during a "random" callout check, Exim treated it as
4609 a failure of the random address, and carried on sending RSET and the real
4610 address. If the delay was just some slowness somewhere, the response to the
4611 original RCPT would be taken as a response to RSET and so on, causing
4612 mayhem of various kinds.
4614 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4615 when I implemented it. It didn't allow for the fact that some option values
4616 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4617 the right test for positive values.
4619 48. Domain names in DNS records are case-independent. Exim always looks them up
4620 in lower case. Some resolvers return domain names in exactly the case they
4621 appear in the zone file, that is, they may contain uppercase letters. Not
4622 all resolvers do this - some return always lower case. Exim was treating a
4623 change of case by a resolver as a change of domain, similar to a widening
4624 of a domain abbreviation. This triggered its re-routing code and so it was
4625 trying to route what was effectively the same domain again. This normally
4626 caused routing to fail (because the router wouldn't handle the domain
4627 twice). Now Exim checks for this case specially, and just changes the
4628 casing of the domain that it ultimately uses when it transmits the message
4631 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4634 50. If a filter generated a file delivery with a non-absolute name (possible if
4635 no home directory exists for the router), the forbid_file option was not
4638 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4639 the existing equality matching.
4641 52. Exim was using ints instead of ino_t variables in some places where it was
4642 dealing with inode numbers.
4644 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4645 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4646 called TMPDIR, and if it finds it is different, it changes its value.
4648 54. The smtp_printf() function is now made available to local_scan() so
4649 additional output lines can be written before returning. There is also an
4650 smtp_fflush() function to enable the detection of a dropped connection.
4651 The variables smtp_input and smtp_batched_input are exported to
4654 55. Changed the default runtime configuration: the message "Unknown user"
4655 has been removed from the ACL, and instead placed on the localuser router,
4656 using the cannot_route_message feature. This means that any verification
4657 failures that generate their own messages won't get overridden. Similarly,
4658 the "Unrouteable address" message that was in the ACL for unverifiable
4659 relay addresses has also been removed.
4661 56. Added hosts_avoid_esmtp to the smtp transport.
4663 57. The exicyclog script was not checking for the esoteric option
4664 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4665 will work only if exicyclog is run under the appropriate euid.
4667 58. Following a discussion on the list, the rules by which Exim recognises line
4668 endings on incoming messages have been changed. The -dropcr and drop_cr
4669 options are now no-ops, retained only for backwards compatibility. The
4670 following line terminators are recognized: LF CRLF CR. However, special
4671 processing applies to CR:
4673 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4674 nor a local message in the state where . is a terminator.
4676 (ii) If a bare CR is encountered in a header line, an extra space is added
4677 after the line terminator so as not to end the header. The reasoning
4678 behind this is that bare CRs in header lines are most likely either
4679 to be mistakes, or people trying to play silly games.
4681 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4682 was being incorrectly given as 18 bytes larger than it should have been.
4683 This is a VOB (very old bug).
4685 60. This may never have affected anything current, but just in case it has:
4686 When the local host is found other than at the start of a list of hosts,
4687 the local host, those with the same MX, and any that follow, are discarded.
4688 When the list in question was part of a longer list of hosts, the following
4689 hosts (not currently being processed) were also being discarded. This no
4690 longer happens. I'm not sure if this situation could ever has previously
4693 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4696 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4697 header. This is something that is recommended in a new Internet Draft, and
4698 is something that is documented as being done by Sendmail. There are two
4699 possible values. For messages generated by the autoreply transport, Exim
4702 Auto-Submitted: auto-replied
4704 whereas for all other generated messages (e.g. bounces) it adds
4706 Auto-Submitted: auto-generated
4708 63. The "personal" condition in filters now includes a test for the
4709 Auto-Submitted: header. If it contains the string "auto-" the message it
4710 not considered personal.
4712 64. Added rcpt_include_affixes as a generic transport option.
4714 65. Added queue_only_override (default true).
4716 66. Added the syslog_duplication option.
4718 67. If what should have been the first header line of a message consisted of
4719 a space followed by a colon, Exim was mis-interpreting it as a header line.
4720 It isn't of course - it is syntactically invalid and should therefore be
4721 treated as the start of the message body. The misbehaviour could have
4722 caused a number of strange effects, including loss of data in subsequent
4723 header lines, and spool format errors.
4725 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4726 client host had authenticated. This control can now be exercised by an ACL
4727 for more flexibility.
4729 69. By default, callouts do not happen when testing with -bh. There is now a
4730 variant, -bhc, which does actually run the callout code, including
4731 consulting and updating the callout cache.
4733 70. Added support for saslauthd authentication, courtesy of Alexander
4736 71. If statvfs() failed on the spool or log directories while checking their
4737 size for availability, Exim confusingly gave the error "space shortage".
4738 Furthermore, in debugging mode it crashed with a floating point exception.
4739 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4740 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4741 serious problem, Exim now writes to the main and panic logs when this
4742 happens, with details of the failure. It then refuses to accept the
4743 incoming message, giving the message "spool directory problem" or "log
4744 directory problem" with a 421 code for SMTP messages.
4746 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4747 0, 1, and 2 exist, because some OS complain for execs without them (see
4748 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4749 descriptors. However, the code omitted to check that the open succeeded,
4750 causing mysterious errors if for some reason the permissions on /dev/null
4751 got screwed. Now Exim writes a message to the main and panic logs, and
4752 bombs out if it can't open /dev/null.
4754 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4755 interact so that it is all more flexible. It is supposed to remain
4756 backwards compatible. Also added extra_local_interfaces.
4758 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4759 to bomb out with an assertion failure - to the client this appears as a
4760 connection drop. This problem occurs in the part of the code that was taken
4761 from the Samba project. Fortunately, the assertion is in a very simple
4762 function, so I have fixed this by reproducing the function inline in the
4763 one place where it is called, and arranging for authentication to fail
4764 instead of killing the process with assert().
4766 75. The SPA client code was not working when the server requested OEM rather
4767 than Unicode encoding.
4769 76. Added code to make require_files with a specific uid setting more usable in
4770 the case where statting the file as root fails - usually a non-root-mounted
4771 NFS file system. When this happens and the failure is EACCES, Exim now
4772 forks a subprocess and does the per-uid checking as the relevant uid.
4774 77. Added process_log_path.
4776 78. If log_file_path was not explicitly set, a setting of check_log_space or
4777 check_log_inodes was ignored.
4779 79. If a space check for the spool or log partitions fails, the incident is now
4780 logged. Of course, in the latter case the data may get lost...
4782 80. Added the %p formatting code to string_format() so that it can be used to
4783 print addresses in debug_print(). Adjusted all the address printing in the
4784 debugging in store.c to use %p rather than %d.
4786 81. There was a concern that a line of code in smtp_in.c could overflow a
4787 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4788 initially expressed, the concern was not well-founded, because trailing
4789 spaces are removed early. However, if the trailing spaces were followed by
4790 a NULL, they did not get removed, so the overflow was possible. Two fixes
4793 (a) I re-wrote the offending code in a cleaner fashion.
4794 (b) If an incoming SMTP command contains a NULL character, it is rejected
4797 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4798 runs initgroups(), so that if the Exim user is in any additional groups,
4799 they will be used during message reception.
4805 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob