1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.517 2007/06/22 14:38:58 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.
76 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
77 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
80 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
83 PH/02 In an ACL statement such as
85 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
87 if a client was not listed at all, or was listed with a value other than
88 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
89 the condition was not true (as it should be), so access was not denied.
90 The bug was that the ! inversion was incorrectly passed on to the second
91 item. This has been fixed.
93 PH/03 Added additional dnslists conditions == and =& which are different from
94 = and & when the dns lookup returns more than one IP address.
96 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
97 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
99 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
100 FSYNC, which compiles an option called disable_fsync that allows for
101 bypassing fsync(). The documentation is heavily laced with warnings.
103 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
105 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
106 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
107 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
108 including adding "make clean"; (3) Added -fPIC when compiling the test
109 dynamically loaded module, to get rid of a warning.
111 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
112 message fails, move_frozen_messages = true and ignore_bounce_errors_after
113 = 0s. The bug is otherwise harmless.
115 PH/07 There was a bug in the dovecot authenticator such that the value of
116 $auth1 could be overwritten, and so not correctly preserved, after a
117 successful authentication. This usually meant that the value preserved by
118 the server_setid option was incorrect.
120 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
122 PH/09 Installed PCRE release 7.0.
124 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
125 run for batched SMTP input. It is now run at the start of every message
126 in the batch. While fixing this I discovered that the process information
127 (output by running exiwhat) was not always getting set for -bs and -bS
128 input. This is fixed, and it now also says "batched" for BSMTP.
130 PH/11 Added control=no_pipelining.
132 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
133 patch, slightly modified), and move the expansion of helo_data till after
134 the connection is made in the smtp transport (so it can use these
137 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
139 PH/14 Added log_selector = +pid.
141 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
143 PH/16 Add ${if forany and ${if forall.
145 PH/17 Added dsn_from option to vary the From: line in DSNs.
147 PH/18 Flush SMTP output before performing a callout, unless control =
148 no_callout_flush is set.
150 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
151 was true (the default) a successful delivery failed to delete the retry
152 item, thus causing premature timeout of the address. The bug is now
155 PH/20 Added hosts_avoid_pipelining to the smtp transport.
157 PH/21 Long custom messages for fakedefer and fakereject are now split up
158 into multiline reponses in the same way that messages for "deny" and
159 other ACL rejections are.
161 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
162 with slight modification.
164 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
165 draft, changing the syntax and factoring some duplicate code".
167 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
168 for deliveries of the second and subsequent messages over the same SMTP
171 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
172 ${reduce, with only minor "tidies".
174 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
176 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
177 expansion side effects.
179 PH/27 When a message times out after an over-quota error from an Exim-imposed
180 quota, the bounce message says "mailbox is full". This message was not
181 being given when it was a system quota that was exceeded. It now should
184 MH/03 Made $recipients available in local_scan(). local_scan() already has
185 better access to the recipient list through recipients_list[], but
186 $recipients can be useful in postmaster-provided expansion strings.
188 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
189 in the case of a MAIL command with additional options following the
190 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
191 were accidentally chopped off.
193 PH/29 SMTP synchronization checks are implemented when a command is read -
194 there is a check that no more input is waiting when there shouldn't be
195 any. However, for some commands, a delay in an ACL can mean that it is
196 some time before the response is written. In this time, more input might
197 arrive, invalidly. So now there are extra checks after an ACL has run for
198 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
199 pipelining has not been advertised.
201 PH/30 MH's patch to allow iscntrl() characters to be list separators.
203 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
204 returned in the SMTP response when smtp_return_error_details was false.
207 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
208 instead of the C I/O that was originally supplied, because problems were
211 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
212 Exim which did not show up earlier: it was assuming that a call to
213 SSL_CTX_set_info_callback() might give an error value. In fact, there is
214 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
215 was a macro that became an assignment, so it seemed to work. This has
216 changed to a proper function call with a void return, hence the compile
217 error. Exim's code has been fixed.
219 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
222 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
224 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
226 PH/37 If a message is not accepted after it has had an id assigned (e.g.
227 because it turns out to be too big or there is a timeout) there is no
228 "Completed" line in the log. When some messages of this type were
229 selected by exigrep, they were listed as "not completed". Others were
230 picked up by some special patterns. I have improved the selection
231 criteria to be more general.
233 PH/38 The host_find_failed option in the manualroute router can now be set
234 to "ignore", to completely ignore a host whose IP address cannot be
235 found. If all hosts are ignored, the behaviour is controlled by the new
236 host_all_ignored option.
238 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
239 homing or because of multiple MX records with /mx) generated more than
240 one IP address, and the following item turned out to be the local host,
241 all the secondary addresses of the first item were incorrectly removed
242 from the list, along with the local host and any following hosts (which
243 is what is supposed to happen).
245 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
246 whoever called Exim into the -H file. In the case of the daemon it was
247 behaving confusingly. When first started, it used values for whoever
248 started the daemon, but after a SIGHUP it used the Exim user (because it
249 calls itself on a restart). I have changed the code so that it now always
252 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
253 message are rejected with the same error (e.g. no authentication or bad
254 sender address), and a DATA command is nevertheless sent (as can happen
255 with PIPELINING or a stupid MUA), the error message that was given to the
256 RCPT commands is included in the rejection of the DATA command. This is
257 intended to be helpful for MUAs that show only the final error to their
260 PH/42 Another patch from the Sieve maintainer.
262 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
263 Eximstats - Fixed some broken HTML links and added missing column headers
265 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
266 columns for Rejects, Temp Rejects, Ham, and Spam rows.
268 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
270 PH/43 Yet another patch from the Sieve maintainer.
272 PH/44 I found a way to check for a TCP/IP connection going away before sending
273 the response to the final '.' that terminates a message, but only in the
274 case where the client has not sent further data following the '.'
275 (unfortunately, this is allowed). However, in many cases there won't be
276 any further data because there won't be any more messages to send. A call
277 to select() can be used: if it shows that the input is "ready", there is
278 either input waiting, or the socket has been closed. An attempt to read
279 the next input character can distinguish the two cases. Previously, Exim
280 would have sent an OK response which the client would never have see.
281 This could lead to message repetition. This fix should cure that, at
282 least in a lot of common cases.
284 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
285 advertised in response to EHLO.
291 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
292 fixed by 4.65/MH/01 (is this a record?) are fixed:
294 (i) An empty string was always treated as zero by the numeric comparison
295 operators. This behaviour has been restored.
297 (ii) It is documented that the numeric comparison operators always treat
298 their arguments as decimal numbers. This was broken in that numbers
299 starting with 0 were being interpreted as octal.
301 While fixing these problems I realized that there was another issue that
302 hadn't been noticed. Values of message_size_limit (both the global option
303 and the transport option) were treated as octal if they started with 0.
304 The documentation was vague. These values are now always treated as
305 decimal, and I will make that clear in the documentation.
311 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
312 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
315 MH/01 Don't check that the operands of numeric comparison operators are
316 integers when their expansion is in "skipping" mode (fixes bug
317 introduced by 4.64-PH/07).
319 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
320 child addresses, Exim now panics and dies. Previously, because the count
321 is held in a short int, deliveries were likely to be lost. As such a
322 large number of recipients for a single message is ridiculous
323 (performance will be very, very poor), I have chosen to impose a limit
324 rather than extend the field.
330 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
331 leftover -K file (the existence of which was triggered by #402).
332 While we were at it, introduced process PID as part of the -K
333 filename. This should rule out race conditions when creating
336 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
337 processing considerably. Previous code took too long for large mails,
338 triggering a timeout which in turn triggers #401.
340 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
341 in the DK code in transports.c. sendfile() is not really portable,
342 hence the _LINUX specificness.
344 TF/01 In the add_headers option to the mail command in an Exim filter,
345 there was a bug that Exim would claim a syntax error in any
346 header after the first one which had an odd number of characters
349 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
350 callout verification, Exim cached a "reject" for the entire domain. This
351 is correct for most verifications, but it is not correct for a recipient
352 verification with use_sender or use_postmaster set, because in that case
353 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
354 case of MAIL FROM:<> rejection from other early rejections (e.g.
355 rejection of HELO). When verifying a recipient using a non-null MAIL
356 address, the cache is ignored if it shows MAIL FROM:<> rejection.
357 Whatever the result of the callout, the value of the domain cache is
358 left unchanged (for any other kind of callout, getting as far as trying
359 RCPT means that the domain itself is ok).
361 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
364 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
365 manifest itself as EPIPE rather than ECONNECT. When tidying away a
366 session, the daemon ignores ECONNECT errors and logs others; it now
367 ignores EPIPE as well.
369 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
370 (quoted-printable decoding).
372 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
373 later the small subsequent patch to fix an introduced bug.
375 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
377 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
379 PH/08 An error is now given if message_size_limit is specified negative.
381 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
382 to be given (somewhat) arbitrary names.
384 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
387 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
388 miscellaneous code fixes
390 PH/10 Added the log_reject_target ACL modifier to specify where to log
393 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
394 hostname. This is wrong, because it relates to the incoming message (and
395 probably the interface on which it is arriving) and not to the outgoing
396 callout (which could be using a different interface). This has been
397 changed to use the value of the helo_data option from the smtp transport
398 instead - this is what is used when a message is actually being sent. If
399 there is no remote transport (possible with a router that sets up host
400 addresses), $smtp_active_hostname is used.
402 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
403 tweaks were necessary in order to get it to work (see also 21 below):
404 (a) The code assumed that strncpy() returns a negative number on buffer
405 overflow, which isn't the case. Replaced with Exim's string_format()
407 (b) There were several signed/unsigned issues. I just did the minimum
408 hacking in of casts. There is scope for a larger refactoring.
409 (c) The code used strcasecmp() which is not a standard C function.
410 Replaced with Exim's strcmpic() function.
411 (d) The code set only $1; it now sets $auth1 as well.
412 (e) A simple test gave the error "authentication client didn't specify
413 service in request". It would seem that Dovecot has changed its
414 interface. Fortunately there's a specification; I followed it and
415 changed what the client sends and it appears to be working now.
417 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
420 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
421 address A is aliased to B and C, where B exists and C does not. Without
422 -v the output is "A verified" because verification stops after a
423 successful redirection if more than one address is generated. However,
424 with -v the child addresses are also verified. Exim was outputting "A
425 failed to verify" and then showing the successful verification for C,
426 with its parentage. It now outputs "B failed to verify", showing B's
427 parentage before showing the successful verification of C.
429 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
430 look up a TXT record in a specific list after matching in a combined
433 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
434 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
435 they consult the DNS. I had assumed they would set it the way they
436 wanted; and indeed my experiments on Linux seem to show that in some
437 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
438 To be on the safe side, however, I have now made the interface to
439 host_find_byname() similar to host_find_bydns(), with an argument
440 containing the DNS resolver options. The host_find_byname() function now
441 sets these options at its start, just as host_find_bydns() does. The smtp
442 transport options dns_qualify_single and dns_search_parents are passed to
443 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
444 of host_find_byname() use the default settings of RES_DEFNAMES
445 (qualify_single) but not RES_DNSRCH (search_parents).
447 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
448 spool_read_header() do less string testing, by means of a preliminary
449 switch on the second character of optional "-foo" lines. (This is
450 overdue, caused by the large number of possibilities that now exist.
451 Originally there were few.) While I was there, I also converted the
452 str(n)cmp tests so they don't re-test the leading "-" and the first
453 character, in the hope this might squeeze out yet more improvement.
455 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
456 flag allowing group syntax was set by the header_syntax check but not
457 turned off, possible causing trouble later; (2) The flag was not being
458 set at all for the header_verify test, causing "group"-style headers to
459 be rejected. I have now set it in this case, and also caused header_
460 verify to ignore an empty address taken from a group. While doing this, I
461 came across some other cases where the code for allowing group syntax
462 while scanning a header line wasn't quite right (mostly, not resetting
463 the flag correctly in the right place). These bugs could have caused
464 trouble for malformed header lines. I hope it is now all correct.
466 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
467 with the "reply" argument non-NULL. The code, however (which originally
468 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
469 but it didn't always do it. This confused somebody who was copying the
470 code for some other use. I have removed all the tests.
472 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
473 feature that was used to support insecure browsers during the U.S. crypto
474 embargo. It requires special client support, and Exim is probably the
475 only MTA that supported it -- and would never use it because real RSA is
476 always available. This code has been removed, because it had the bad
477 effect of slowing Exim down by computing (never used) parameters for the
478 RSA_EXPORT functionality.
480 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
481 authenticator to fail if there's a tab character in the incoming data
482 (there should never be unless someone is messing about, as it's supposed
483 to be base64-encoded). Also added, on Timo's advice, the "secured" option
484 if the connection is using TLS or if the remote IP is the same as the
485 local IP, and the "valid-client-cert option" if a client certificate has
488 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
489 authenticators. This can be used for authorization after authentication
490 succeeds. (In the case of plaintext, it servers for both authentication
493 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
494 if any retry times were supplied.
496 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
497 connection, where there is no client IP address to check. In this
498 situation, the verify now always succeeds.
500 PH/25 Applied John Jetmore's -Mset patch.
502 PH/26 Added -bem to be like -Mset, but loading a message from a file.
504 PH/27 In a string expansion for a processed (not raw) header when multiple
505 headers of the same name were present, leading whitespace was being
506 removed from all of them, but trailing whitespace was being removed only
507 from the last one. Now trailing whitespace is removed from each header
508 before concatenation. Completely empty headers in a concatenation (as
511 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
512 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
514 PH/29 [Removed. This was a change that I later backed out, and forgot to
515 correct the ChangeLog entry (that I had efficiently created) before
516 committing the later change.]
518 PH/30 Exim was sometimes attempting to deliver messages that had suffered
519 address errors (4xx response to RCPT) over the same connection as other
520 messages routed to the same hosts. Such deliveries are always "forced",
521 so retry times are not inspected. This resulted in far too many retries
522 for the affected addresses. The effect occurred only when there were more
523 hosts than the hosts_max_try setting in the smtp transport when it had
524 the 4xx errors. Those hosts that it had tried were not added to the list
525 of hosts for which the message was waiting, so if all were tried, there
526 was no problem. Two fixes have been applied:
528 (i) If there are any address or message errors in an SMTP delivery, none
529 of the hosts (tried or untried) are now added to the list of hosts
530 for which the message is waiting, so the message should not be a
531 candidate for sending over the same connection that was used for a
532 successful delivery of some other message. This seems entirely
533 reasonable: after all the message is NOT "waiting for some host".
534 This is so "obvious" that I'm not sure why it wasn't done
535 previously. Hope I haven't missed anything, but it can't do any
536 harm, as the worst effect is to miss an optimization.
538 (ii) If, despite (i), such a delivery is accidentally attempted, the
539 routing retry time is respected, so at least it doesn't keep
540 hammering the server.
542 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
543 in ${readsocket because some servers need this prod.
545 PH/32 Added some extra debug output when updating a wait-xxx database.
547 PH/33 The hint "could be header name not terminated by colon", which has been
548 given for certain expansion errors for a long time, was not being given
549 for the ${if def:h_colon_omitted{... case.
551 PH/34 The spec says: "With one important exception, whenever a domain list is
552 being scanned, $domain contains the subject domain." There was at least
553 one case where this was not true.
555 PH/35 The error "getsockname() failed: connection reset by peer" was being
556 written to the panic log as well as the main log, but it isn't really
557 panic-worthy as it just means the connection died rather early on. I have
558 removed the panic log writing for the ECONNRESET error when getsockname()
561 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
562 runs only) independently of the message's sender address. This meant
563 that, if the 4xx error was in fact related to the sender, a different
564 message to the same recipient with a different sender could confuse
565 things. In particualar, this can happen when sending to a greylisting
566 server, but other circumstances could also provoke similar problems.
567 I have changed the default so that the retry time for these errors is now
568 based a combination of the sender and recipient addresses. This change
569 can be overridden by setting address_retry_include_sender=false in the
572 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
573 remote server are returned as part of bounce messages. This was not
574 happening for LMTP over a pipe (the lmtp transport), but now it is the
575 same for both kinds of LMTP.
577 PH/38 Despite being documented as not happening, Exim was rewriting addresses
578 in header lines that were in fact CNAMEs. This is no longer the case.
580 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
581 and queue runs started by the daemon processed all messages. This has
582 been fixed so that -R and -S can now usefully be given with -q<time>.
584 PH/40 Import PCRE release 6.7 (fixes some bugs).
586 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
588 PH/42 Give an error if -q is specified more than once.
590 PH/43 Renamed the variables $interface_address and $interface_port as
591 $received_ip_address and $received_port, to make it clear that these
592 values apply to message reception, and not to the outgoing interface when
593 a message is delivered. (The old names remain recognized, of course.)
595 PH/44 There was no timeout on the connect() call when using a Unix domain
596 socket in the ${readsocket expansion. There now is.
598 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
599 be meaningful with "accept".
601 SC/01 Eximstats V1.43
602 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
604 SC/02 Eximstats V1.44
605 Use a glob alias rather than an array ref in the generated
606 parser. This improves both readability and performance.
608 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
609 Collect SpamAssassin and rejection statistics.
610 Don't display local sender or destination tables unless
611 there is data to show.
612 Added average volumes into the top table text output.
614 SC/04 Eximstats V1.46
615 Collect data on the number of addresses (recipients)
616 as well as the number of messages.
618 SC/05 Eximstats V1.47
619 Added 'Message too big' to the list of mail rejection
620 reasons (thanks to Marco Gaiarin).
622 SC/06 Eximstats V1.48
623 Mainlog lines which have GMT offsets and are too short to
624 have a flag are now skipped.
626 SC/07 Eximstats V1.49 (Alain Williams)
627 Added the -emptyok flag.
629 SC/08 Eximstats V1.50
630 Fixes for obtaining the IP address from reject messages.
632 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
633 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
634 whitesspace changes from 4.64-PH/27
636 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
639 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
640 are found, allow negative numbers in numeric criteria)
642 JJ/06 exipick.20061117.2, added new $message_body_missing variable
644 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
645 to match changes made in 4.64-PH/43
647 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
649 PH/47 Put in an explicit test for a DNS lookup of an address record where the
650 "domain" is actually an IP address, and force a failure. This locks out
651 those revolvers/nameservers that support "A-for-A" lookups, in
652 contravention of the specifications.
654 PH/48 When a host name was looked up from an IP address, and the subsequent
655 forward lookup of the name timed out, the host name was left in
656 $sender_host_name, contrary to the specification.
658 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
659 restricted to single-key lookups, Exim was not diagnosing an error if
660 * or *@ was used with a query-style lookup.
662 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
664 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
665 long ago, but noone interested enough thought of it. Let's just say that
666 the "1.1" means that there are some new functions that weren't there at
667 some point in the past.
669 PH/51 Error processing for expansion failure of helo_data from an smtp
670 transport during callout processing was broken.
672 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
673 tested/used via the -bh/-bhc/-bs options.
675 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
676 bug, fixed in subsequent PCRE releases).
678 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
679 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
681 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
687 SC/01 Use a glob alias rather than an array ref in eximstats generated
688 parser. This improves both readability and performance.
690 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
691 Don't display local sender or destination tables in eximstats unless
692 there is data to show.
693 Added average volumes into the eximstats top table text output.
695 SC/03 Collect data on the number of addresses (recipients) as well
696 as the number of messages in eximstats.
698 TF/01 Correct an error in the documentation for the redirect router. Exim
699 does (usually) call initgroups() when daemonizing.
701 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
702 with consistent privilege compared to when running as a daemon.
704 TF/03 Note in the spec that $authenticated_id is not set for local
705 submissions from trusted users.
707 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
708 Thanks to Dean Brooks <dean@iglou.com> for the patch.
710 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
711 by adding some example configuration directives to the default
712 configuration file. A little bit of work is required to uncomment the
713 directives and define how usernames and passwords are checked, but
714 there is now a framework to start from.
716 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
717 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
718 without this. I don't know how relevant this is to other LDAP libraries.
720 PH/02 Add the verb name to the "unknown ACL verb" error.
722 PH/03 Magnus Holmgren's patch for filter_prepend_home.
724 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
726 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
727 directory not expanded when it should be if an expanded home directory
728 was set for the address (which is overridden by the transport).
730 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
733 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
734 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
735 because it is too late at that time, and has no effect.
737 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
738 security issue with \' (bugzilla #107). I could not use the
739 PQescapeStringConn() function, because it needs a PGconn value as one of
742 PH/08 When testing addresses using -bt, indicate those final addresses that
743 are duplicates that would not cause an additional delivery. At least one
744 person was confused, thinking that -bt output corresponded to deliveries.
745 (Suppressing duplicates isn't a good idea as you lose the information
746 about possibly different redirections that led to the duplicates.)
748 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
749 systems where poll() doesn't work, in particular OS X.
751 PH/10 Added more information to debugging output for retry time not reached.
753 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
754 operations in malware.c.
756 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
759 PH/13 If write_rejectlog was set false when logging was sent to syslog with
760 syslog_duplication set false, log lines that would normally be written
761 both the the main log and to the reject log were not written to syslog at
764 PH/14 In the default configuration, change the use of "message" in ACL warn
765 statements to "add_header".
767 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
768 not followed by a command (e.g. "seen endif").
770 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
771 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
774 PH/17 Added extra conditions to the default value of delay_warning_condition
778 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
779 { match{$h_precedence:}{(?i)bulk|list|junk} } \
780 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
783 The Auto-Submitted: and various List- headers are standardised, whereas I
784 don't think Precedence: ever was.
786 PH/18 Refactored debugging code in route_finduser() to show more information,
787 in particular, the error code if getpwnam() issues one.
789 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
790 This is apparently needed in addition to the PH/07 change above to avoid
791 any possible encoding problems.
793 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
794 but not after initializing Perl.
796 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
797 output them only if debugging. By default they are written stderr,
798 apparently, which is not desirable.
800 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
803 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
806 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
808 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
809 authenticated or an ident call has been made. Suppress the default
810 values for $authenticated_id and $authenticated_sender (but permit -oMai
811 and -oMas) when testing with -bh.
813 PH/24 Re-jigged the order of the tests in the default configuration so that the
814 tests for valid domains and recipients precede the DNS black list and CSA
815 tests, on the grounds that those ones are more expensive.
817 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
818 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
819 This bug exists in every version of Exim that I still have, right back to
822 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
823 However, an attempt to turn on case-sensitivity in a regex key by
824 including (?-i) didn't work because the subject string was already
825 lowercased, and the effects were non-intuitive. It turns out that a
826 one-line patch can be used to allow (?-i) to work as expected.
832 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
833 other effects) broke the use of negated acl sub-conditions.
835 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
838 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
839 "Deny" causes Exim to reject the incoming connection with a 554 error.
840 Unfortunately, if there is a major crisis, such as a disk failure,
841 tcp-wrappers gives "deny", whereas what one would like would be some
842 kind of temporary error. A kludge has been added to help with this.
843 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
844 554 error is used if errno is still zero or contains ENOENT (which occurs
845 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
848 PH/03 Add -lutil to the default FreeBSD LIBS setting.
850 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
851 errors. Otherwise a message that provokes a temporary error (when other
852 messages do not) can cause a whole host to time out.
854 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
855 addresses were routed directly to files or pipes from a redirect router.
856 File deliveries just didn't batch; pipe deliveries might have suffered
859 PH/06 A failure to get a lock for a hints database would erroneously always say
860 "Failed to get write lock", even when it was really a read lock.
862 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
863 of 0600. This has been changed to use the value of the lockfile_mode
864 option (which defaults to 0600).
866 PH/08 Applied small patch from the Sieve maintainer.
868 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
869 folder from quota calculations, a direct delivery into this folder messed
870 up the contents of the maildirsize file. This was because the regex was
871 used only to exclude .Trash (or whatever) when the size of the mailbox
872 was calculated. There was no check that a delivery was happening into an
873 excluded directory. This bug has been fixed by ignoring all quota
874 processing for deliveries into excluded directories.
876 PH/10 Added the maildirfolder_create_regex option to appendfile.
882 PH/01 The code for finding all the local interface addresses on a FreeBSD
883 system running IPv6 was broken. This may well have applied to all BSD
884 systems, as well as to others that have similar system calls. The broken
885 code found IPv4 interfaces correctly, but gave incorrect values for the
886 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
887 that it would not match correctly against @[] and not recognize the IPv6
890 PH/02 The ipliteral router was not recognizing addresses of the form user@
891 [ipv6:....] because it didn't know about the "ipv6:" prefix.
893 PH/03 Added disable_ipv6.
895 PH/04 Changed $reply_address to use the raw form of the headers instead of the
896 decoded form, because it is most often used to construct To: headers
897 lines in autoreplies, and the decoded form may well be syntactically
898 invalid. However, $reply_address has leading white space removed, and all
899 newlines turned into spaces so that the autoreply transport does not
902 PH/05 If group was specified without a user on a router, and no group or user
903 was specified on a transport, the group from the router was ignored.
905 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
906 for visible compile-time settings that can be used to change these
907 numbers, for those that want even more. Backwards compatibility with old
908 spool files has been maintained. However, going back to a previous Exim
909 release will lost any variables that are in spool files.
911 PH/07 Two small changes when running in the test harness: increase delay when
912 passing a TCP/IP connection to a new process, in case the original
913 process has to generate a bounce, and remove special handling of
914 127.0.0.2 (sic), which is no longer necessary.
916 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
917 be the same on different OS.
919 PH/09 Moved a debug statement in filter processing to avoid a race problem when
922 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
923 whether --show-vars was specified or not
925 JJ/02 exipick: Added support for new ACL variable spool format introduced
928 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
929 syntactically invalid From: or Reply-to: line, and a filter used this to
930 generate an autoreply, and therefore failed to obtain an address for the
931 autoreply, Exim could try to deliver to a non-existent relative file
932 name, causing unrelated and misleading errors. What now happens is that
933 it logs this as a hard delivery error, but does not attempt to create a
936 PH/11 The exinext utility has a -C option for testing purposes, but although
937 the given file was scanned by exinext itself; it wasn't being passed on
938 when Exim was called.
940 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
941 an end-of-file indication when reading a command response.
943 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
944 compiled. In many other places in Exim, IPv6 addresses are always
945 recognized, so I have changed this. It also means that IPv4 domain
946 literals of the form [IPV4:n.n.n.n] are now always recognized.
948 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
949 used if the router is not running as root, for example, when verifying at
950 ACL time, or when using -bh. The debugging output from this situation was
951 non-existent - all you got was a failure to exec. I have made two
954 (a) Failures to set uid/gid, the current directory, or a process leader
955 in a subprocess such as that created by queryprogram now generate
956 suitable debugging ouput when -d is set.
958 (b) The queryprogram router detects when it is not running as root,
959 outputs suitable debugging information if -d is set, and then runs
960 the subprocess without attempting to change uid/gid.
962 PH/15 Minor change to Makefile for building test_host (undocumented testing
965 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
966 additional section of a DNS packet that returns MX or SRV records.
967 Instead, it always explicitly searches for A/AAAA records. This avoids
968 major problems that occur when a DNS server includes only records of one
969 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
970 fixed another bug: if SRV records were looked up and the corresponding
971 address records were *not* found in the additional section, the port
972 values from the SRV records were lost.
974 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
975 using the correct key (the original address) when searching the retry
976 rules in order to find which one to use for generating the retry hint.
978 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
979 adding the default one. Similarly, if it contains a Reply-To: header, the
980 errors_reply_to option, if set, is not used.
982 PH/19 When calculating a retry time, Exim used to measure the "time since
983 failure" by looking at the "first failed" field in the retry record. Now
984 it does not use this if it is later than than the arrival time of the
985 message. Instead it uses the arrival time. This makes for better
986 behaviour in cases where some deliveries succeed, thus re-setting the
987 "first failed" field. An example is a quota failure for a huge message
988 when small messages continue to be delivered. Without this change, the
989 "time since failure" will always be short, possible causing more frequent
990 delivery attempts for the huge message than are intended.
991 [Note: This change was subsequently modified - see PH/04 for 4.62.]
993 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
994 $1, $2, $3) because the numerical variables can be reset during some
995 expansion items (e.g. "match"), thereby losing the authentication data.
997 PH/21 Make -bV show the size of off_t variables so that the test suite can
998 decide whether to run tests for quotas > 2G.
1000 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1001 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1002 filecount value is greater than 2G or if a quota value is greater than 2G
1003 on a system where the size of off_t is not greater than 4, a panic error
1006 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1007 never match. The debug and -bh output now contains an explicit error
1008 message indicating a malformed IPv4 address or mask.
1010 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1011 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1012 PH/23 above applies.
1014 PH/25 Do not write to syslog when running in the test harness. The only
1015 occasion when this arises is a failure to open the main or panic logs
1016 (for which there is an explicit test).
1018 PH/26 Added the /no_tell option to "control=freeze".
1020 PH/27 If a host name lookup failed very early in a connection, for example, if
1021 the IP address matched host_lookup and the reverse lookup yielded a name
1022 that did not have a forward lookup, an error message of the form "no IP
1023 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1024 could be logged. Now it outputs the IP address instead of "NULL".
1026 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1027 allows the sender and the authenticated sender to be set when
1028 submitting a message from within Exim. Since child_open_exim() is
1029 documented for local_scan(), the new function should be too.
1031 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1032 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1033 results in an empty string is now treated as unset.
1035 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1037 PH/31 Added sender_verify_fail logging option.
1039 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1040 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1041 tidied the source and removed it altogether.
1043 PH/33 When a queue run was abandoned because the load average was too high, a
1044 log line was always written; now it is written only if the queue_run log
1045 selector is set. In addition, the log line for abandonment now contains
1046 information about the queue run such as the pid. This is always present
1047 in "start" and "stop" lines but was omitted from the "abandon" line.
1049 PH/34 Omit spaces between a header name and the colon in the error message that
1050 is given when verify = headers_syntax fails (if there are lots of them,
1051 the message gets confusing).
1053 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1054 names, as there are now some PTR records that contain slashes. This check
1055 is only to protect against broken name servers that fall over on strange
1056 characters, so the fact that it applies to all lookups doesn't matter.
1058 PH/36 Now that the new test suite is complete, we can remove some of the
1059 special code in Exim that was needed for the old test suite. For example,
1060 sorting DNS records because real resolvers return them in an arbitrary
1061 order. The new test suite's fake resolver always returns records in the
1064 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1065 bounces) except when queue_only is set, to avoid logging races between
1066 the different processes.
1068 PH/38 Panic-die if .include specifies a non-absolute path.
1070 PH/39 A tweak to the "H" retry rule from its user.
1072 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1073 a label. They prevented compilation on older perls.
1075 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1076 a warning to be raised on newish perls.
1078 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1079 on queue. Changes to match documented behaviour of showing count of
1080 messages matching specified criteria.
1082 PH/40 Changed the default ident timeout from 30s to 5s.
1084 PH/41 Added support for the use of login_cap features, on those BSD systems
1085 that have them, for controlling the resources used by pipe deliveries.
1087 PH/42 The content-scanning code uses fopen() to create files in which to put
1088 message data. Previously it was not paying any attention to the mode of
1089 the files. Exim runs with umask(0) because the rest of the code creates
1090 files with open(), and sets the required mode explicitly. Thus, these
1091 files were ending up world-writeable. This was not a big issue, because,
1092 being within the spool directory, they were not world-accessible. I have
1093 created a function called modefopen, which takes an additional mode
1094 argument. It sets umask(777), creates the file, chmods it to the required
1095 mode, then resets the umask. All the relevant calls to fopen() in the
1096 content scanning code have been changed to use this function.
1098 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1099 to 24 hours. This avoids potential overflow problems when processing G
1100 and H retry rules. I suspect nobody ever tinkers with this value.
1102 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1104 PH/45 When the plaintext authenticator is running as a client, the server's
1105 challenges are checked to ensure they are valid base64 strings. By
1106 default, the authentication attempt is cancelled if an invalid string is
1107 received. Setting client_ignore_invalid_base64 true ignores these errors.
1108 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1109 they are received. Thus, the responses can be made to depend on the
1110 challenges. If an invalid string is ignored, an empty string is placed in
1113 PH/46 Messages that are created by the autoreply transport now contains a
1114 References: header, in accordance with RFCs 2822 and 3834.
1116 PH/47 Added authenticated_sender_force to the smtp transport.
1118 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1120 PH/49 Installed latest patch from the Sieve maintainer.
1122 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1123 was also set, the appendfile transport was unnecessarily scanning a
1124 directory of message files (e.g. for maildir delivery) to find the count
1125 of files (along with the size), even though it did not need this
1126 information. It now does the scan only if it needs to find either the
1127 size of the count of files.
1129 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1131 PH/52 Two bugs concerned with error handling when the smtp transport is
1134 (i) Exim was not creating retry information for temporary errors given
1135 for individual recipients after the DATA command when the smtp transport
1136 was used in LMTP mode. This meant that they could be retried too
1137 frequently, and not timed out correctly.
1139 (ii) Exim was setting the flag that allows error details to be returned
1140 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1141 recipients that were returned after the DATA command.
1143 PH/53 This is related to PH/52, but is more general: for any failing address,
1144 when detailed error information was permitted to be returned to the
1145 sender, but the error was temporary, then after the final timeout, only
1146 "retry timeout exceeded" was returned. Now it returns the full error as
1147 well as "retry timeout exceeded".
1149 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1150 do this, and (what is worse) MTAs that accept it.
1152 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1153 will now be deprecated.
1155 PH/56 New os.c-cygwin from the Cygwin maintainer.
1157 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1158 formats (previously only available in exim formats via -bpr, -bpru,
1159 and -bpra. Now also available in native and exiqgrep formats)
1161 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1162 with very large, slow to parse queues
1164 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1166 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1168 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1169 responses to authentication challenges, though it was showing the
1170 challenges; (ii) I've removed the CR characters from the debug output for
1173 PH/58 Allow for the insertion of a newline as well as a space when a string
1174 is turned into more than one encoded-word during RFC 2047 encoding. The
1175 Sieve code now uses this.
1177 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1178 data_4xx, lost_connection, tls_required.
1180 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1181 message was being sent as an SMTP response.
1183 PH/61 Add -l and -k options to exicyclog.
1185 PH/62 When verifying, if an address was redirected to one new address, so that
1186 verification continued, and the new address failed or deferred after
1187 having set something in $address_data, the value of $address_data was not
1188 passed back to the ACL. This was different to the case when no
1189 redirection occurred. The value is now passed back in both cases.
1191 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1192 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1193 use login_cap.h, so on its own it isn't the distinguishing feature. The
1194 new name refers directly to the setclassresources() function.
1196 PH/65 Added configuration files for NetBSD3.
1198 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1200 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1201 is preferred over IPv4.
1203 PH/68 The bounce_return_message and bounce_return_body options were not being
1204 honoured for bounces generated during the reception of non-SMTP messages.
1205 In particular, this applied to messages rejected by the ACL. This bug has
1206 been fixed. However, if bounce_return_message is true and bounce_return_
1207 body is false, the headers that are returned for a non-SMTP message
1208 include only those that have been read before the error was detected.
1209 (In the case of an ACL rejection, they have all been read.)
1211 PH/69 The HTML version of the specification is now built in a directory called
1212 spec_html instead of spec.html, because the latter looks like a path with
1213 a MIME-type, and this confuses some software.
1215 PH/70 Catch two compiler warnings in sieve.c.
1217 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1218 function verify_get_ident() calls ip_connect() to connect a socket, but
1219 if the "connect()" function timed out, ip_connect() used to close the
1220 socket. However, verify_get_ident() also closes the socket later, and in
1221 between Exim writes to the log, which may get opened at this point. When
1222 the socket was closed in ip_connect(), the log could get the same file
1223 descriptor number as the socket. This naturally causes chaos. The fix is
1224 not to close the socket in ip_connect(); the socket should be closed by
1225 the function that creates it. There was only one place in the code where
1226 this was missing, in the iplookup router, which I don't think anybody now
1227 uses, but I've fixed it anyway.
1229 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1230 well as to direct DNS lookups. Otherwise the handling of names in host
1231 lists is inconsistent and therefore confusing.
1237 PH/01 Two changes to the default runtime configuration:
1239 (1) Move the checks for relay_from_hosts and authenticated clients from
1240 after to before the (commented out) DNS black list checks.
1242 (2) Add control=submission to the relay_from_hosts and authenticated
1243 clients checks, on the grounds that messages accepted by these
1244 statements are most likely to be submissions.
1246 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1248 (1) Generate an error if the third argument for the ${prvs expansion is
1251 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1254 (3) Reset the variables that are obtained from the first argument of
1255 ${prvscheck and used in the second argument before leaving the code,
1256 because their memory is reclaimed, so using them afterwards may do
1259 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1260 one (it's much easier than Tom thought :-).
1262 (5) Because of (4), we can now allow for the use of $prvscheck_result
1263 inside the third argument.
1265 PH/03 For some reason, the default setting of PATH when running a command from
1266 a pipe transport was just "/usr/bin". I have changed it to
1269 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1270 anything to be listed in the output from -bV.
1272 PH/05 When a filter generated an autoreply, the entire To: header line was
1273 quoted in the delivery log line, like this:
1275 => >A.N.Other <ano@some.domain> <original@ddress> ...
1277 This has been changed so that it extracts the operative address. There
1278 may be more than one such address. If so, they are comma-separated, like
1281 => >ano@some.domain,ona@other.domain <original@ddress> ...
1283 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1284 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1285 not being looked up in the rDNS to get a host name, Exim was showing the
1286 IP address twice in Received: lines, even though the IP addresses were
1287 identical. For example:
1289 Received: from [1.2.3.4] (helo=[1.2.3.4])
1291 However, if the real host name was known, it was omitting the HELO data
1292 if it matched the actual IP address. This has been tidied up so that it
1293 doesn't show the same IP address twice.
1295 PH/07 When both +timestamp and +memory debugging was on, the value given by
1296 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1297 called by the debug printing, thereby overwriting the timestamp buffer.
1298 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1300 PH/08 When the original message was included in an autoreply transport, it
1301 always said "this is a copy of the message, including all the headers",
1302 even if body_only or headers_only was set. It now gives an appropriate
1305 PH/09 Applied a patch from the Sieve maintainer which:
1307 o fixes some comments
1308 o adds the (disabled) notify extension core
1309 o adds some debug output for the result of if/elsif tests
1310 o points to the current vacation draft in the documentation
1311 and documents the missing references header update
1315 o fixes a bug in processing the envelope test (when testing
1316 multiple envelope elements, the last element determinted the
1319 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1320 Electronic Mail") by including:
1322 Auto-submitted: auto-generated
1324 in the messages that it generates (bounce messages and others, such as
1325 warnings). In the case of bounce messages for non-SMTP mesages, there was
1326 also a typo: it was using "Auto_submitted" (underscore instead of
1327 hyphen). Since every message generated by Exim is necessarily in response
1328 to another message, thes have all been changed to:
1330 Auto-Submitted: auto-replied
1332 in accordance with these statements in the RFC:
1334 The auto-replied keyword:
1336 - SHOULD be used on messages sent in direct response to another
1337 message by an automatic process,
1339 - MUST NOT be used on manually-generated messages,
1341 - MAY be used on Delivery Status Notifications (DSNs) and Message
1342 Disposition Notifications (MDNs),
1344 - MUST NOT be used on messages generated by automatic or periodic
1345 processes, except for messages which are automatic responses to
1348 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1349 to the default Received: header definition.
1351 PH/12 Added log selector acl_warn_skipped (default on).
1353 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1354 variables because (a) they are in the wrong storage pool and (b) even if
1355 they were copied, it wouldn't work properly because of the caching.
1357 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1358 checking when decoding. Apparently there are clients that generate
1359 overlong encoded strings. Why am I not surprised?
1361 PH/15 If the first argument of "${if match_address" was not empty, but did not
1362 contain an "@" character, Exim crashed. Now it writes a panic log message
1363 and treats the condition as false.
1365 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1367 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1368 extension "envelope-auth". The code is finished and in agreement with
1369 other implementations, but there is no documentation so far and in fact,
1370 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1371 not changing the active code.
1373 Print executed "if" and "elsif" statements when debugging is used. This
1374 helps a great deal to understand what a filter does.
1376 Document more things not specified clearly in RFC3028. I had all this
1377 sorted out, when out of a sudden new issues came to my mind. Oops."
1379 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1382 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1383 canonical form (as documented). However, after a host name lookup from
1384 the IP address, check_host() was doing a simple string comparison with
1385 addresses acquired from the DNS when checking that the found name did
1386 have the original IP as one of its addresses. Since any found IPv6
1387 addresses are likely to be in abbreviated form, the comparison could
1388 fail. Luckily, there already exists a function for doing the comparison
1389 by converting both addresses to binary, so now that is used instead of
1390 the text comparison.
1392 PH/20 There was another similar case to PH/19, when a complete host name was
1393 given in a host list; looking up its IP address could give an abbreviated
1394 form, whereas the current host's name might or might not be abbreviated.
1395 The same fix has been applied.
1401 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1402 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1405 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1406 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1408 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1410 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1411 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1412 Also, for "Auto-Submitted", treat anything other than "no" as
1413 non-personal, in accordance with RFC 3834. (Previously it treated
1414 anything starting "auto-" as non-personal.)
1416 TF/01 The control=submission/name=... option had a problem with syntax
1417 errors if the name included a slash character. The /name= option
1418 now slurps the rest of the string, so it can include any characters
1419 but it must come last in the list of options (after /sender_retain
1422 PH/05 Some modifications to the interface to the fake nameserver for the new
1430 TK/01 Added the "success_on_redirect" address verification option. See
1431 NewStuff for rationale and an example.
1433 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1435 PH/02 Patch to exigrep to allow it to work on syslog lines.
1437 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1438 fread() to skip over the body file's header line, because in Cygwin the
1439 header line is locked and is inaccessible.
1441 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1442 co-exist for some time) to make it clear that it is the Exim ID that is
1443 referenced, not the Message-ID: header line.
1445 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1446 string_format() function, because snprintf() does not exist on all
1449 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1450 ${stat: expansion item.
1452 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1453 protocol synchronization error", to keep the pedants happy.
1455 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1456 well as for IRIX systems, when gcc is being used. See the host.c source
1459 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1461 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1464 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1465 transport and to the smtp transport in LMTP mode.
1467 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1469 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1470 run a filter in a subprocess. This could lead to confusion in subsequent
1471 lookups in the parent process. There should also be a search_tidyup() at
1472 the end of the subprocess.
1474 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1475 only if the host matched helo_try_verify_hosts, which caused the
1476 verification to occur when the EHLO/HELO command was issued. The ACL just
1477 tested the remembered result. Now, if a previous verification attempt has
1478 not happened, "verify = helo" does it there and then.
1480 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1482 TK/03 Fix log output including CR from clamd.
1484 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1485 exist provoked a memory error which could cause a segfault.
1487 PH/15 Installed PCRE 6.2
1489 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1491 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1492 of the problem. Specifically, suggested +O2 rather than +O1 for the
1495 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1497 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1498 record, the port was not correctly logged when the outgoing_port log
1499 selector was set (it logged the transort's default port).
1501 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1502 fallback_hosts, and "hosts" in the smtp transport.
1504 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1505 host errors such as "Connection refused".
1507 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1508 authentication with radiusclient 0.4.9:
1510 - Error returned from rc_read_config was caught wrongly
1511 - Username/password not passed on to radius server due to wrong length.
1513 The presumption is that some radiusclient API changes for 4.51/PH/17
1514 were not taken care of correctly. The code is still untested by me (my
1515 Linux distribution still has 0.3.2 of radiusclient), but it was
1516 contributed by a Radius user.
1518 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1519 expanding the "port" option of the smtp transport.
1521 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1522 while reading a MIME header. Thanks to Tom Hughes for a patch.
1524 PH/24 Include config.h inside local_scan.h so that configuration settings are
1527 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1528 that in an ACL for RCPT (for example), you can examine exactly what was
1531 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1532 commands, but it was not correctly comparing the address with the actual
1533 client host address. Thus, it would show the EHLO address in Received:
1534 header lines when this was not necessary.
1536 PH/27 Added the % operator to ${eval:}.
1538 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1539 it should be ignoring failures (because with -C, for example, it has lost
1540 privilege). It wasn't ignoring creation failures other than "already
1543 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1546 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1547 because an input line was too long, either on its own, or by virtue of
1548 too many continuations, the temporary file was not being removed, and the
1549 return code was incorrect.
1551 PH/31 Missing "BOOL" in function definition in filtertest.c.
1553 PH/32 Applied Sieve patches from the maintainer.
1555 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1557 PH/33 Added "verify = not_blind".
1559 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1560 Local/Makefile (with some defaults set). These are used in built scripts
1561 such as exicyclog, but they have never been used in the exim_install
1562 script (though there are many overriding facilities there). I have
1563 arranged that the exim_install script now takes note of these two
1566 PH/35 Installed configuration files for Dragonfly.
1568 PH/36 When a locally submitted message by a trusted user did not contain a
1569 From: header, and the sender address was obtained from -f or from an SMTP
1570 MAIL command, and the trusted user did not use -F to supply a sender
1571 name, $originator_name was incorrectly used when constructing a From:
1572 header. Furthermore, $originator_name was used for submission mode
1573 messages from external hosts without From: headers in a similar way,
1574 which is clearly wrong.
1576 PH/37 Added control=suppress_local_fixups.
1578 PH/38 When log_selector = +received_sender was set, and the addition of the
1579 sender made the log line's construction buffer exactly full, or one byte
1580 less than full, an overflow happened when the terminating "\n" was
1583 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1584 when the result of a list match is failure because a DNS lookup failed.
1586 PH/40 RM_COMMAND is now used in the building process.
1588 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1589 the "build-* directories that it finds.
1591 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1592 address in a domain literal was a prefix of an interface address.
1594 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1595 when verifying a sender address, unless rewrite_headers is false.
1597 PH/44 Wrote a long comment about why errors_to addresses are verified as
1598 recipients, not senders.
1600 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1601 the ratelimit ACL was added.
1603 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1605 PH/46 Added extra information about PostgreSQL errors to the error string.
1607 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1608 avoiding the need to install special zones in a real server. This is
1609 backwards compatible; if it can't find the fake resolver, it drops back.
1610 Thus, both old and new test suites can be run.
1612 TF/02 Added util/ratelimit.pl
1614 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1615 clock is set back in time.
1617 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1618 Candler <B.Candler@pobox.com>.
1620 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1621 OK for addresses that are the result of redirections.
1623 PH/48 A number of further additions for the benefit of the new test suite,
1624 including a fake gethostbyname() that interfaces to the fake DNS resolver
1627 TF/06 The fix for widen_domains has also been applied to qualify_single and
1628 search_parents which are the other dnslookup options that can cause
1631 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1634 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1636 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1637 service" if it encountered an MX record with an empty target hostname.
1638 The message is now "an MX or SRV record indicated no SMTP service".
1640 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1641 if the DNS of the sending site is misconfigured. This is quite a
1642 common situation. This change restores the behaviour of treating a
1643 helo verification defer as a failure.
1645 PH/51 If self=fail was set on a router, the bounce message did not include the
1646 actual error message.
1652 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1654 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1655 command itself ended in error, the underlying message about the transport
1656 filter timeout was being overwritten with the pipe command error. Now the
1657 underlying error message should be appended to the second error message.
1659 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1661 PH/02 Reduce the amount of output that "make" produces by default. Full output
1662 can still be requested.
1664 PH/03 The warning log line about a condition test deferring for a "warn" verb
1665 was being output only once per connection, rather than after each
1666 occurrence (because it was using the same function as for successful
1667 "warn" verbs). This seems wrong, so I have changed it.
1669 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1670 it should not have, which might have caused a crash in the right
1671 circumstances, but probably never did.
1673 PH/04 Installed a modified version of Tony Finch's patch to make submission
1674 mode fix the return path as well as the Sender: header line, and to
1675 add a /name= option so that you can make the user's friendly name appear
1678 TF/03 Added the control = fakedefer ACL modifier.
1680 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1681 Mark Lowes for thorough testing.
1683 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1685 TK/03 Merged latest SRS patch from Miles Wilton.
1687 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1688 with the definition in sysexits.h (which is #included earlier).
1689 Fortunately, Exim does not actually use EX_OK. The code used to try to
1690 preserve the sysexits.h value, by assumimg that macro definitions were
1691 scanned for macro replacements. I have been disabused of this notion,
1692 so now the code just undefines EX_OK before #including unistd.h.
1694 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1695 in the smtp transport. When a block could not be written in a single
1696 write() function, the timeout was being re-applied to each part-write.
1697 This seems wrong - if the receiver was accepting one byte at a time it
1698 would take for ever. The timeout is now adjusted when this happens. It
1699 doesn't have to be particularly precise.
1701 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1702 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1704 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1705 without a domain if the check to <postmaster@domain> fails.
1707 SC/01 Eximstats: added -xls and the ability to specify output files
1708 (patch written by Frank Heydlauf).
1710 SC/02 Eximstats: use FileHandles for outputing results.
1712 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1714 SC/04 Eximstats: fixed display of large numbers with -nvr option
1716 SC/05 Eximstats: fixed merging of reports with empty tables.
1718 SC/06 Eximstats: added the -include_original_destination flag
1720 SC/07 Eximstats: removed tabs and trailing whitespace.
1722 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1724 TK/06 MBOX spool code: Add real "From " MBOX separator line
1725 so the .eml file is really in mbox format (even though
1726 most programs do not really care). Patch from Alex Miller.
1728 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1729 The latter is generated from $received_to and is only set if the
1730 message has one envelope recipient. SA can use these headers,
1731 obviously out-of-the-box. Patch from Alex Miller.
1733 PH/08 The ${def test on a variable was returning false if the variable's
1734 value was "0", contrary to what the specification has always said!
1735 The result should be true unless the variable is empty.
1737 PH/09 The syntax error of a character other than { following "${if
1738 def:variable_name" (after optional whitespace) was not being diagnosed.
1739 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1740 accidental colon was present, for example, could give incorrect results.
1742 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1743 result is used (not including appendfile, where other changes are about
1746 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1747 This involved changing a lot of size variables from int to off_t. It
1748 should work with maildirs and everything.
1750 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1751 spamd dying while we are connected to it.
1753 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1756 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1757 function for BDB 4.3.
1759 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1761 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1762 the sources for PCRE have been re-arranged and more files are now
1765 PH/15 The code I had for printing potentially long long variables in PH/11
1766 above was not the best (it lost precision). The length of off_t variables
1767 is now inspected at build time, and an appropriate printing format (%ld
1768 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1769 to be "long long int" or "long int". This is needed for the internal
1770 formatting function string_vformat().
1772 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1773 the configuration file to be ":syslog", then the script "guesses" where
1774 the logs files are, rather than using the compiled in default. In our
1775 case the guess is not the same as the compiled default, so the script
1776 suddenly stopped working when I started to use syslog. The patch checks
1777 to see if log_file_path is "". If so, it attempts to read it from exim
1778 with no configuration file to get the compiled in version, before it
1779 falls back to the previous guessing code."
1781 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1782 implementing BATV in an Exim configuration. See NewStuff for the gory
1785 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1786 Makefile that are specific to HP-UX.
1788 PH/18 If the "use_postmaster" option was set for a recipient callout together
1789 with the "random" option, the postmaster address was used as the MAIL
1790 FROM address for the random test, but not for the subsequent recipient
1791 test. It is now used for both.
1793 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1794 patch removes a few documentation additions to RFC 3028, because the
1795 latest draft now contains them. It adds the new en;ascii-case comparator
1796 and a new error check for 8bit text in MIME parts. Comparator and
1797 require names are now matched exactly. I enabled the subaddress
1798 extension, but it is not well tested yet (read: it works for me)."
1800 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1801 rework some of the code of TK/09 above to avoid the hardwired use of
1802 "%lld" and "long long". Replaced the call to snprintf() with a call to
1805 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1806 records point to non-existent hosts", "retry timeout exceeded", and
1807 "retry time not reached for any host after a long failure period".
1809 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1810 experimental DomainKeys support:
1812 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1813 (2) On an error such as an illegally used "control", the wrong name for
1814 the control was given.
1816 These problems did NOT occur unless DomainKeys support was compiled.
1818 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1820 PH/24 Added ${if match_ip condition.
1822 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1823 databases so that it will be absolutely obvious if a crash occurs in the
1824 DB library. This is a regular occurrence (often caused by mis-matched
1827 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1828 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1829 picked up on a user's system that detects such things. There doesn't seem
1830 to be a gcc warning option for this - only an attribute that has to be
1831 put on the function's prototype. It seems that in Fedora Core 4 they have
1832 set this on a number of new functions. No doubt there will be more in due
1835 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1836 specify a transport. However, if an address that was verified by such a
1837 router was the subject of a callout, Exim crashed because it tried to
1838 read the rcpt_include_affixes from the non-existent transport. Now it
1839 just assumes that the setting of that option is false. This bug was
1840 introduced by 4.51/PH/31.
1842 PH/28 Changed -d+all to exclude +memory, because that information is very
1843 rarely of interest, but it makes the output a lot bigger. People tend to
1844 do -d+all out of habit.
1846 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1847 code in os-type was giving problems when libc.so lives in lib64, like on
1850 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1851 aren't the modern standard, and it seems that some systems' include files
1852 don't always have them. Exim was already checking for some of the newer
1853 ones like T_AAAA, and defining it itself. I've added checks for all the
1854 record types that Exim uses.
1856 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1857 not automatically generating a new one, as it is supposed to. This
1858 prevented TLS from working. If the file did exist, but contained invalid
1859 data, a new version was generated, as expected. It was only the case of a
1860 non-existent file that was broken.
1862 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1863 with a change in libdomainkeys > 0.64.
1865 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1866 from DNS. If the selector record carries the flag, it now has
1867 precedence over the domain-wide flag.
1869 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1871 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1872 the use of a port name that isn't defined in /etc/services) occurred, the
1873 message was deferred as in a normal delivery, and thus remained on the
1874 spool, instead of being failed because of the mua_wrapper setting. This
1875 is now fixed, and I tidied up some of the mua_wrapper messages at the
1878 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1879 the messages in a hash of arrays rather than using individual hashes.
1880 This is a bit cleaner and results in dramatic memory savings, albeit
1881 at a slight CPU cost.
1883 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1884 as requested by Marc Sherman.
1886 SC/10 Eximstats: added histograms for user specified patterns as requested
1889 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1891 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1892 fopen() in the content-scanning modules that did not already have it.
1898 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1899 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1901 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1903 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1905 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1906 to test Sieve filters that use "vacation".
1908 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1909 that changes the way the GnuTLS parameters are stored in the cache file.
1910 The new format can be generated externally. For backward compatibility,
1911 if the data in the cache doesn't make sense, Exim assumes it has read an
1912 old-format file, and it generates new data and writes a new file. This
1913 means that you can't go back to an older release without removing the
1916 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1917 work if there are any delivery delays because "one_time" forces the
1918 parent to be marked "delivered", so its unseen clone is never tried
1919 again. For this reason, Exim now forbids the simultaneous setting of
1922 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1923 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1924 entry if you want to know the details. The fix, however, neglected to
1925 consider the case when local delivery batching is involved. The test for
1926 "previously delivered" was not happening when checking to see if an
1927 address could be batched with a previous (undelivered) one; under
1928 certain circumstances this could lead to multiple deliveries to the same
1931 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1932 in its include files, and this causes problems building Exim.
1934 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1935 header_syntax) but Exim was just ignoring anything given after a slash.
1936 In particular, this caused confusion with an attempt to use "verify =
1937 reverse_host_lookup/defer_ok". An error is now given when options are
1938 supplied for verify items that do not have them. (Maybe reverse_host_
1939 lookup should have a defer_ok option, but that's a different point.)
1941 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1942 defined by RFC 821) to 2048, because there were problems with some AUTH
1943 commands, and RFC 1869 says the size should be increased for extended
1944 SMTP commands that take arguments.
1946 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1949 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1950 "unknown" error; now it says that the functionality isn't in the binary.
1952 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1953 an address' error message when a string expansion fails (syntax or
1954 whatever). Otherwise the password may appear in the log. Following change
1955 PH/42 below, there is no longer a chance of it appearing in a bounce
1958 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1960 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1961 of its addresses was ever tried. (Bugzilla bug #2.)
1963 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1964 the result incorrectly in the debug output. (It correctly added a newline
1965 to what was transported.)
1967 TF/01 Added $received_time.
1969 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1970 commented out examples of how to interface to a virus scanner and to
1971 SpamAssassin. Also added commented examples of av_scanner and
1972 spamd_address settings.
1974 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1975 and controls are allowed in which ACLs. There were a couple of minor
1976 errors. Some of the entries in the conditions table (which is a table of
1977 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1978 negation of where the condition IS allowed.
1980 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1982 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1983 header file does not have a version number, so I've had to invent a new
1984 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1985 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1986 radiusclient), but it was contributed by a Radius user.
1988 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1989 files or directories, for OpenSSL.
1991 PH/19 When an Exim process that is running as root has to create an Exim log
1992 file, it does so in a subprocess that runs as exim:exim so as to get the
1993 ownership right at creation (otherwise, other Exim processes might see
1994 the file with the wrong ownership). There was no test for failure of this
1995 fork() call, which would lead to the process getting stuck as it waited
1996 for a non-existent subprocess. Forks do occasionally fail when resources
1997 run out. I reviewed all the other calls to fork(); they all seem to check
2000 PH/20 When checking for unexpected SMTP input at connect time (before writing
2001 the banner), Exim was not dealing correctly with a non-positive return
2002 from the read() function. If the client had disconnected by this time,
2003 the result was a log entry for a synchronization error with an empty
2004 string after "input=" when read() returned zero. If read() returned -1
2005 (an event I could not check), uninitialized data bytes were printed.
2006 There were reports of junk text (parts of files, etc) appearing after
2009 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2011 PH/22 Added support for macro redefinition, and (re)definition in between
2012 driver and ACL definitions.
2014 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2015 forgetting to use the resulting value; it was using the unexpanded value.
2017 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2018 hadn't been configured. The fix is from Juergen Kreileder, who
2019 understands it better than I do:
2021 "Here's what I see happening with three configured cyrus_sasl
2022 authenticators configured (plain, login, cram-md5):
2024 On startup auth_cyrus_sasl_init() gets called for each of these.
2025 This means three calls to sasl_listmech() without a specified mech_list.
2026 => SASL tests which mechs of all available mechs actually work
2027 => three warnings about OTP not working
2028 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2030 With the patch, sasl_listmech() also gets called three times. But now
2031 SASL's mech_list option is set to the server_mech specified in the the
2032 authenticator. Or in other words, the answer from sasl_listmech()
2033 gets limited to just the mech you're testing for (which is different
2035 => the return list contains just 'plain' or 'login', 'cram-md5' or
2036 nothing depending on the value of ob->server_mech.
2038 I've just tested the patch: Authentication still works fine,
2039 unavailable mechs specified in the exim configuration are still
2040 caught, and the auth.log warnings about OTP are gone."
2042 PH/25 When debugging is enabled, the contents of the command line are added
2043 to the debugging output, even when log_selector=+arguments is not
2046 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2047 answer is "GNU", and only if the return is "GNU/something" is the answer
2050 PH/27 $acl_verify_message is now set immediately after the failure of a
2051 verification in an ACL, and so is available in subsequent modifiers. In
2052 particular, the message can be preserved by coding like this:
2054 warn !verify = sender
2055 set acl_m0 = $acl_verify_message
2057 Previously, $acl_verify_message was set only while expanding "message"
2058 and "log_message" when a very denied access.
2060 PH/28 Modified OS/os.c-Linux with
2062 -#ifndef OS_LOAD_AVERAGE
2063 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2065 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2066 nomenclature these days.)
2068 PH/29 Installed patch from the Sieve maintainer that adds the options
2069 sieve_useraddress and sieve_subaddress to the redirect router.
2071 PH/30 In these circumstances:
2072 . Two addresses routed to the same list of hosts;
2073 . First host does not offer TLS;
2074 . First host accepts first address;
2075 . First host gives temporary error to second address;
2076 . Second host offers TLS and a TLS session is established;
2077 . Second host accepts second address.
2078 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2079 and peerdn, if requested) that were in fact used only for the second
2082 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2083 attention to any local part prefix or suffix that was matched by the
2084 router that accepted the address. It now behaves in the same way as it
2085 does for delivery: the affixes are removed from the local part unless
2086 rcpt_include_affixes is set on the transport.
2088 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2089 timeout during the DATA phase of an incoming message.
2091 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2092 applied a patch sanctioned by the Sieve maintainer.
2094 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2095 the uid or gid is negative. A case of a negative gid caused this to be
2096 noticed. The fix allows for either to be negative.
2098 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2099 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2102 PH/36 The change PH/12 above was broken. Fixed it.
2104 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2105 the grounds that routing the same address twice would always produce the
2106 same answer. This might have been true once, but it is certainly no
2107 longer true now. Routing a child address may depend on the previous
2108 routing that produced that child. Some complicated redirection strategies
2109 went wrong when messages had multiple recipients, and made Exim's
2110 behaviour dependent on the order in which the addresses were given.
2112 I have moved the duplicate checking until after the routing is complete.
2113 Exim scans the addresses that are assigned to local and remote
2114 transports, and removes any duplicates. This means that more work will be
2115 done, as duplicates will always all be routed, but duplicates are
2116 presumably rare, so I don't expect this is of any significance.
2118 For deliveries to pipes, files, and autoreplies, the duplicate checking
2119 still happens during the routing process, since they are not going to be
2122 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2123 It corrects a timeout issue with spamd. This is Ian's comment: "The
2124 background is that sometimes spamd either never reads data from a
2125 connection it has accepted, or it never writes response data. The exiscan
2126 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2127 blindly assumes that writes won't block so it may never time out."
2129 PH/39 Allow G after quota size as well as K and M.
2131 PH/40 The value set for $authenticated_id in an authenticator may not contain
2132 binary zeroes or newlines because the value is written to log lines and
2133 to spool files. There was no check on this. Now the value is run through
2134 the string_printing() function so that such characters are converted to
2135 printable escape sequences.
2137 PH/41 $message_linecount is a new variable that contains the total number of
2138 lines in the message. Compare $body_linecount, which is the count for the
2141 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2142 bounce and delay warning messages, except in certain special cases, which
2145 (a) An SMTP error message from a remote host;
2146 (b) A message specified in a :fail: redirection;
2147 (c) A message specified in a "fail" command in a system filter;
2148 (d) A message specified in a FAIL return from the queryprogram router;
2149 (e) A message specified by the cannot_route_message router option.
2151 In these cases only, Exim does include the error details in bounce and
2152 warning messages. There are also a few cases where bland messages such
2153 as "unrouteable address" or "local delivery error" are given.
2155 PH/43 $value is now also set for the "else" part of a ${run expansion.
2157 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2158 being worked on, but at least Exim now implements the latest version to
2161 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2162 process to complete was treated as a delivery failure, a timeout while
2163 writing the message to the pipe was logged, but erroneously treated as a
2164 successful delivery. Such timeouts include transport filter timeouts. For
2165 consistency with the overall process timeout, these timeouts are now
2166 treated as errors, giving rise to delivery failures by default. However,
2167 there is now a new Boolean option for the pipe transport called
2168 timeout_defer, which, if set TRUE, converts the failures into defers for
2169 both kinds of timeout. A transport filter timeout is now identified in
2172 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2173 systems where "make" and "gmake" are different, calling "gmake" at top
2174 level broke things. I've arranged for the value of $(MAKE) to be passed
2175 from the Makefile to this script so that it can call the same version of
2179 A note about Exim versions 4.44 and 4.50
2180 ----------------------------------------
2182 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2183 changes of various kinds. As a consequence, a big documentation update was
2184 needed. This delayed the release for rather longer than seemed good, especially
2185 in the light of a couple of (minor) security issues. Therefore, the changes
2186 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2187 release. So 4.44 and 4.50 are in effect two different branches that both start
2190 I have left the 4.50 change log unchanged; it contains all the changes since
2191 4.43. The change log for 4.44 is below; many of its items are identical to
2192 those for 4.50. This seems to be the most sensible way to preserve the
2193 historical information.
2199 1. Minor wording change to the doc/README.SIEVE file.
2201 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2202 computation of the current number of files was incorrect.
2204 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2205 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2208 4. Give more explanation in the error message when the command for a transport
2209 filter fails to execute.
2211 5. There are several places where Exim runs a non-Exim command in a
2212 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2213 was being done only for the command run by the queryprogram router. It is
2214 now done for all such subprocesses. The other cases are: ${run, transport
2215 filters, and the commands run by the lmtp and pipe transports.
2217 6. Added CONFIGURE_GROUP build-time option.
2219 7. Some older OS have a limit of 256 on the maximum number of file
2220 descriptors. Exim was using setrlimit() to set 1000 as a large value
2221 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2222 systems. I've change it so that if it can't get 1000, it tries for 256.
2224 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2225 was an oversight, and furthermore, ever since the addition of extra
2226 controls (e.g. 4.43/32), the checks on when to allow different forms of
2227 "control" were broken. There should now be diagnostics for all cases when a
2228 control that does not make sense is encountered.
2230 9. Added the /retain_sender option to "control=submission".
2232 10. $recipients is now available in the predata ACL (oversight).
2234 11. Tidy the search cache before the fork to do a delivery from a message
2235 received from the command line. Otherwise the child will trigger a lookup
2236 failure and thereby defer the delivery if it tries to use (for example) a
2237 cached ldap connection that the parent has called unbind on.
2239 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2240 of $address_data from the recipient verification was clobbered by the
2241 sender verification.
2243 13. The value of address_data from a sender verification is now available in
2244 $sender_address_data in subsequent conditions in the ACL statement.
2246 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2248 15. Added a new option "connect=<time>" to callout options, to set a different
2251 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2252 was its contents. (It was OK if the option was not defined at all.)
2254 17. A "Completed" log line is now written for messages that are removed from
2255 the spool by the -Mrm option.
2257 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2258 information about exactly what failed.
2260 19. Added -dd to debug only the daemon process.
2262 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2263 handles timeouts, both on the server side and network timeouts. Renamed the
2264 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2266 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2267 It is now set to "smtps".
2269 22. $host_address is now set to the target address during the checking of
2270 ignore_target_hosts.
2272 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2273 being passed; this would have caused $sender_host_name to have been used if
2274 matching the list had actually called for a host name (not very likely,
2275 since this list is usually IP addresses). A host name is now passed as
2278 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2279 code that specifies a non-restarting handler (typically sigaction() in
2280 modern systems) in an attempt to fix a rare and obscure crash bug.
2282 25. Narrowed the window for a race in the daemon that could cause it to ignore
2283 SIGCHLD signals. This is not a major problem, because they are used only to
2284 wake it up if nothing else does.
2286 26. A malformed maildirsize file could cause Exim to calculate negative values
2287 for the mailbox size or file count. Odd effects could occur as a result.
2288 The maildirsize information is now recalculated if the size or filecount
2291 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2292 support for a long time. Removed HAVE_SYS_VFS_H.
2294 28. Installed the latest version of exipick from John Jetmore.
2296 29. In an address list, if the pattern was not a regular expression, an empty
2297 subject address (from a bounce message) matched only if the pattern was an
2298 empty string. Non-empty patterns were not even tested. This was the wrong
2299 because it is perfectly reasonable to use an empty address as part of a
2300 database query. An empty address is now tested by patterns that are
2301 lookups. However, all the other forms of pattern expect the subject to
2302 contain a local part and a domain, and therefore, for them, an empty
2303 address still always fails if the pattern is not itself empty.
2305 30. Exim went into a mad DNS loop when attempting to do a callout where the
2306 host was specified on an smtp transport, and looking it up yielded more
2307 than one IP address.
2309 31. Re-factored the code for checking spool and log partition space into a
2310 function that finds that data and another that does the check. The former
2311 is then used to implement four new variables: $spool_space, $log_space,
2312 $spool_inodes, and $log_inodes.
2314 32. The RFC2047 encoding function was originally intended for short strings
2315 such as real names; it was not keeping to the 75-character limit for
2316 encoded words that the RFC imposes. It now respects the limit, and
2317 generates multiple encoded words if necessary. To be on the safe side, I
2318 have increased the buffer size for the ${rfc2047: expansion operator from
2321 33. It is now permitted to omit both strings after an "if" condition; if the
2322 condition is true, the result is "true". As before, when the second string
2323 is omitted, a false condition yields an empty string. This makes it less
2324 cumbersome to write custom ACL and router conditions.
2326 34. Failure to deliver a bounce message always caused it to be frozen, even if
2327 there was an errors_to setting on the router. The errors_to setting is now
2330 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2331 canonical form (fully expanded) before being placed in
2332 $sender_host_address.
2334 36. The table in the code that translates DNS record types into text (T_A to
2335 "A" for instance) was missing entries for NS and CNAME. It is just possible
2336 that this could have caused confusion if both these types were looked up
2337 for the same domain, because the text type is used as part of Exim's
2338 per-process caching. But the chance of anyone hitting this buglet seems
2341 37. The dnsdb lookup has been extended in a number of ways.
2343 (1) There is a new type, "zns", which walks up the domain tree until it
2344 finds some nameserver records. It should be used with care.
2346 (2) There is a new type, "mxh", which is like "mx" except that it returns
2347 just the host names, not the priorities.
2349 (3) It is now possible to give a list of domains (or IP addresses) to be
2350 looked up. The behaviour when one of the lookups defers can be
2351 controlled by a keyword.
2353 (4) It is now possible to specify the separator character for use when
2354 multiple records are returned.
2356 38. The dnslists ACL condition has been extended: it is now possible to supply
2357 a list of IP addresses and/or domains to be looked up in a particular DNS
2360 39. Added log_selector=+queue_time_overall.
2362 40. When running the queue in the test harness, wait just a tad after forking a
2363 delivery process, to get repeatability of debugging output.
2365 41. Include certificate and key file names in error message when GnuTLS fails
2366 to set them up, because the GnuTLS error message doesn't include the name
2367 of the failing file when there is a problem reading it.
2369 42. Allow both -bf and -bF in the same test run.
2371 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2373 44. The "Exiscan patch" is now merged into the mainline Exim source.
2375 45. Sometimes the final signoff response after QUIT could fail to get
2376 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2377 < 0 before doing a fflush(). This bug looks as though it goes back to the
2378 introduction of TLS in release 3.20, but "sometimes" must have been rare
2379 because the tests only now provoked it.
2381 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2382 (this can affect the format of dates).
2384 47. exim_tidydb, when checking for the continued existence of a message for
2385 which it has found a message-specific retry record, was not finding
2386 messages that were in split spool directories. Consequently, it was
2387 deleting retry records that should have stayed in existence.
2389 48. Steve fixed some bugs in eximstats.
2391 49. The SPA authentication driver was not abandoning authentication and moving
2392 on to the next authenticator when an expansion was forced to fail,
2393 contradicting the general specification for all authenticators. Instead it
2394 was generating a temporary error. It now behaves as specified.
2396 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2397 (the order specifies the preference for clients). The order is now AES256,
2398 AES128, 3DES, ARCFOUR128.
2400 51. Small patch to Sieve code - explicitly set From: when generating an
2403 52. Exim crashed if a remote delivery caused a very long error message to be
2404 recorded - for instance if somebody sent an entire SpamAssassin report back
2405 as a large number of 550 error lines. This bug was coincidentally fixed by
2406 increasing the size of one of Exim's internal buffers (big_buffer) that
2407 happened as part of the Exiscan merge. However, to be on the safe side, I
2408 have made the code more robust (and fixed the comments that describe what
2411 53. Now that there can be additional text after "Completed" in log lines (if
2412 the queue_time_overall log selector is set), a one-byte patch to exigrep
2413 was needed to allow it to recognize "Completed" as not the last thing in
2416 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2417 patch that reportedly fixes this has been added. I am not expert enough to
2418 create a test for it. This is what the patch creator wrote:
2420 "I found a little strange behaviour of ldap code when working with
2421 Windows 2003 AD Domain, where users was placed in more than one
2422 Organization Units. When I tried to give exim partial DN, the exit code
2423 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2424 But simultaneously result of request was absolutely normal ldap result,
2425 so I produce this patch..."
2427 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2428 REFERENCE, so I have modified the code to exclude the patch when that macro
2431 55. Some experimental protocols are using DNS PTR records for new purposes. The
2432 keys for these records are domain names, not reversed IP addresses. The
2433 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2434 leaves it alone. Component reversal etc. now happens only for IP addresses.
2437 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2439 57. Double the size of the debug message buffer (to 2048) so that more of very
2440 long debug lines gets shown.
2442 58. The exicyclog utility now does better if the number of log files to keep
2443 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2445 59. Two changes related to the smtp_active_hostname option:
2447 (1) $smtp_active_hostname is now available as a variable.
2448 (2) The default for smtp_banner uses $smtp_active_hostname instead
2449 of $primary_hostname.
2451 60. The host_aton() function is supposed to be passed a string that is known
2452 to be a valid IP address. However, in the case of IPv6 addresses, it was
2453 not checking this. This is a hostage to fortune. Exim now panics and dies
2454 if the condition is not met. A case was found where this could be provoked
2455 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2456 components; fortuitously, this particular loophole had already been fixed
2457 by change 4.50/55 above.
2459 If there are any other similar loopholes, the new check in host_aton()
2460 itself should stop them being exploited. The report I received stated that
2461 data on the command line could provoke the exploit when Exim was running as
2462 exim, but did not say which command line option was involved. All I could
2463 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2464 running as the user.
2467 61. There was a buffer overflow vulnerability in the SPA authentication code
2468 (which came originally from the Samba project). I have added a test to the
2469 spa_base64_to_bits() function which I hope fixes it.
2472 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2473 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2475 63. The daemon start-up calls getloadavg() while still root for those OS that
2476 need the first call to be done as root, but it missed one case: when
2477 deliver_queue_load_max is set with deliver_drop_privilege. This is
2478 necessary for the benefit of the queue runner, because there is no re-exec
2479 when deliver_drop_privilege is set.
2481 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2482 This has been fixed.
2484 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2485 was in use, was not putting the data itself into the right store pool;
2486 consequently, it could be overwritten for a subsequent message in the same
2487 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2490 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2492 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2493 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2494 the function were treating the return as a boolean value, which happened to
2495 work because 0=false and not-0=true, but is not correct code.
2497 68. The host_aton() function was not handling scoped IPv6 addresses (those
2498 with, for example, "%eth0" on the end) correctly.
2500 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2501 negated items (that is, ~something) in unsigned ints. Some compilers
2502 apparently mutter when there is no cast.
2504 70. If an address verification called from an ACL failed, and did not produce a
2505 user-specific message (i.e. there was only a "system" message), nothing was
2506 put in $acl_verify_message. In this situation, it now puts the system
2509 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2510 session; change 4.31/43 added the unwanted input to the log line - except
2511 that it did not do this in the start of session case. It now does.
2513 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2514 This is wrong and can cause the other end to generate a synchronization
2515 error if it is another Exim or anything else that does the synchronization
2516 check. A QUIT command is no longer sent after a timeout.
2518 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2519 during host lookups.
2521 74. The defer_ok option of callout verification was not working if it was used
2522 when verifying addresses in header lines, that is, for this case:
2524 verify = header_sender/callout=defer_ok
2526 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2527 those file descriptors could be used for SMTP connections. If anything
2528 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2529 could be sent to the SMTP client, causing chaos. The daemon now opens
2530 stdin, stdout, and stderr to /dev/null when it puts itself into the
2533 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2534 log by default. The user can override this with suitable Perl magic.
2536 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2537 the log message when discard triggers, was not working for the DATA ACL or
2538 for the non-SMTP ACL.
2540 78. Error message wording change in sieve.c.
2542 79. If smtp_accept_max_per_host was set, the number of connections could be
2543 restricted to fewer than expected, because the daemon was trying to set up
2544 a new connection before checking whether the processes handling previous
2545 connections had finished. The check for completed processes is now done
2546 earlier. On busy systems, this bug wouldn't be noticed because something
2547 else would have woken the daemon, and it would have reaped the completed
2550 80. If a message was submitted locally by a user whose login name contained one
2551 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2552 It caused a spool format error. I have fixed the spool reading code. A
2553 related problem was that the "from" clause in the Received: line became
2554 illegal because of the space(s). It is now covered by ${quote_local_part.
2556 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2559 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2561 83. Patch from Sieve maintainer to fix unterminated string problem in
2562 "vacation" handling.
2564 84. Some minor changes to the Linux configuration files to help with other
2565 OS variants using glibc.
2567 85. One more patch for Sieve to update vacation handling to latest spec.
2570 ----------------------------------------------------
2571 See the note above about the 4.44 and 4.50 releases.
2572 ----------------------------------------------------
2578 1. Change 4.43/35 introduced a bug that caused file counts to be
2579 incorrectly computed when quota_filecount was set in an appendfile
2582 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2583 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2586 3. Give more explanation in the error message when the command for a transport
2587 filter fails to execute.
2589 4. There are several places where Exim runs a non-Exim command in a
2590 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2591 was being done only for the command run by the queryprogram router. It is
2592 now done for all such subprocesses. The other cases are: ${run, transport
2593 filters, and the commands run by the lmtp and pipe transports.
2595 5. Some older OS have a limit of 256 on the maximum number of file
2596 descriptors. Exim was using setrlimit() to set 1000 as a large value
2597 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2598 systems. I've change it so that if it can't get 1000, it tries for 256.
2600 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2601 was an oversight, and furthermore, ever since the addition of extra
2602 controls (e.g. 4.43/32), the checks on when to allow different forms of
2603 "control" were broken. There should now be diagnostics for all cases when a
2604 control that does not make sense is encountered.
2606 7. $recipients is now available in the predata ACL (oversight).
2608 8. Tidy the search cache before the fork to do a delivery from a message
2609 received from the command line. Otherwise the child will trigger a lookup
2610 failure and thereby defer the delivery if it tries to use (for example) a
2611 cached ldap connection that the parent has called unbind on.
2613 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2614 of $address_data from the recipient verification was clobbered by the
2615 sender verification.
2617 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2618 was its contents. (It was OK if the option was not defined at all.)
2620 11. A "Completed" log line is now written for messages that are removed from
2621 the spool by the -Mrm option.
2623 12. $host_address is now set to the target address during the checking of
2624 ignore_target_hosts.
2626 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2627 being passed; this would have caused $sender_host_name to have been used if
2628 matching the list had actually called for a host name (not very likely,
2629 since this list is usually IP addresses). A host name is now passed as
2632 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2633 code that specifies a non-restarting handler (typically sigaction() in
2634 modern systems) in an attempt to fix a rare and obscure crash bug.
2636 15. Narrowed the window for a race in the daemon that could cause it to ignore
2637 SIGCHLD signals. This is not a major problem, because they are used only to
2638 wake it up if nothing else does.
2640 16. A malformed maildirsize file could cause Exim to calculate negative values
2641 for the mailbox size or file count. Odd effects could occur as a result.
2642 The maildirsize information is now recalculated if the size or filecount
2645 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2646 support for a long time. Removed HAVE_SYS_VFS_H.
2648 18. Updated exipick to current release from John Jetmore.
2650 19. Allow an empty sender to be matched against a lookup in an address list.
2651 Previously the only cases considered were a regular expression, or an
2654 20. Exim went into a mad DNS lookup loop when doing a callout where the
2655 host was specified on the transport, if the DNS lookup yielded more than
2658 21. The RFC2047 encoding function was originally intended for short strings
2659 such as real names; it was not keeping to the 75-character limit for
2660 encoded words that the RFC imposes. It now respects the limit, and
2661 generates multiple encoded words if necessary. To be on the safe side, I
2662 have increased the buffer size for the ${rfc2047: expansion operator from
2665 22. Failure to deliver a bounce message always caused it to be frozen, even if
2666 there was an errors_to setting on the router. The errors_to setting is now
2669 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2670 canonical form (fully expanded) before being placed in
2671 $sender_host_address.
2673 24. Updated eximstats to version 1.33
2675 25. Include certificate and key file names in error message when GnuTLS fails
2676 to set them up, because the GnuTLS error message doesn't include the name
2677 of the failing file when there is a problem reading it.
2679 26. Expand error message when OpenSSL has problems setting up cert/key files.
2682 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2683 (this can affect the format of dates).
2685 28. exim_tidydb, when checking for the continued existence of a message for
2686 which it has found a message-specific retry record, was not finding
2687 messages that were in split spool directories. Consequently, it was
2688 deleting retry records that should have stayed in existence.
2690 29. eximstats updated to version 1.35
2691 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2692 1.35 - bugfix such that pie charts by volume are generated correctly
2694 30. The SPA authentication driver was not abandoning authentication and moving
2695 on to the next authenticator when an expansion was forced to fail,
2696 contradicting the general specification for all authenticators. Instead it
2697 was generating a temporary error. It now behaves as specified.
2699 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2700 (the order specifies the preference for clients). The order is now AES256,
2701 AES128, 3DES, ARCFOUR128.
2703 31. Small patch to Sieve code - explicitly set From: when generating an
2706 32. Exim crashed if a remote delivery caused a very long error message to be
2707 recorded - for instance if somebody sent an entire SpamAssassin report back
2708 as a large number of 550 error lines. This bug was coincidentally fixed by
2709 increasing the size of one of Exim's internal buffers (big_buffer) that
2710 happened as part of the Exiscan merge. However, to be on the safe side, I
2711 have made the code more robust (and fixed the comments that describe what
2714 33. Some experimental protocols are using DNS PTR records for new purposes. The
2715 keys for these records are domain names, not reversed IP addresses. The
2716 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2717 leaves it alone. Component reversal etc. now happens only for IP addresses.
2720 34. The host_aton() function is supposed to be passed a string that is known
2721 to be a valid IP address. However, in the case of IPv6 addresses, it was
2722 not checking this. This is a hostage to fortune. Exim now panics and dies
2723 if the condition is not met. A case was found where this could be provoked
2724 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2725 components; fortuitously, this particular loophole had already been fixed
2726 by change 4.50/55 or 4.44/33 above.
2728 If there are any other similar loopholes, the new check in host_aton()
2729 itself should stop them being exploited. The report I received stated that
2730 data on the command line could provoke the exploit when Exim was running as
2731 exim, but did not say which command line option was involved. All I could
2732 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2733 running as the user.
2736 35. There was a buffer overflow vulnerability in the SPA authentication code
2737 (which came originally from the Samba project). I have added a test to the
2738 spa_base64_to_bits() function which I hope fixes it.
2741 36. The daemon start-up calls getloadavg() while still root for those OS that
2742 need the first call to be done as root, but it missed one case: when
2743 deliver_queue_load_max is set with deliver_drop_privilege. This is
2744 necessary for the benefit of the queue runner, because there is no re-exec
2745 when deliver_drop_privilege is set.
2747 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2748 was in use, was not putting the data itself into the right store pool;
2749 consequently, it could be overwritten for a subsequent message in the same
2750 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2753 38. Sometimes the final signoff response after QUIT could fail to get
2754 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2755 < 0 before doing a fflush(). This bug looks as though it goes back to the
2756 introduction of TLS in release 3.20, but "sometimes" must have been rare
2757 because the tests only now provoked it.
2763 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2764 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2765 Now it returns an integer. A number of calls were still expecting a T/F
2766 return. Fortuitously, in all cases, the tests worked in OK situations,
2767 which is the norm. However, things would have gone wrong on any write
2768 failures on the smtp file descriptor. This function is used when sending
2769 messages over SMTP and also when doing verify callouts.
2771 2. When Exim is called to do synchronous delivery of a locally submitted
2772 message (the -odf or -odi options), it no longer closes stderr before doing
2775 3. Implemented the mua_wrapper option.
2777 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2779 5. Implemented the functions header_remove(), header_testname(),
2780 header_add_at_position(), and receive_remove_recipient(), and exported them
2783 6. If an ACL "warn" statement specified the addition of headers, Exim already
2784 inserted X-ACL-Warn: at the start if there was no header name. However, it
2785 was not making this test for the second and subsequent header lines if
2786 there were newlines in the string. This meant that an invalid header could
2787 be inserted if Exim was badly configured.
2789 7. Allow an ACL "warn" statement to add header lines at the start or after all
2790 the Received: headers, as well as at the end.
2792 8. Added the rcpt_4xx retry error code.
2794 9. Added postmaster_mailfrom=xxx to callout verification option.
2796 10. Added mailfrom=xxxx to the callout verification option, for verify=
2799 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2800 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2802 12. Inserted some casts to stop certain compilers warning when using pointer
2803 differences as field lengths or precisions in printf-type calls (mostly
2804 affecting debugging statements).
2806 13. Added optional readline() support for -be (dynamically loaded).
2808 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2809 same clock tick as a message's arrival, so that its received time was the
2810 same as the "first fail" time on the retry record, and that message
2811 remained on the queue past the ultimate address timeout, every queue runner
2812 would try a delivery (because it was past the ultimate address timeout) but
2813 after another failure, the ultimate address timeout, which should have then
2814 bounced the address, did not kick in. This was a "< instead of <=" error;
2815 in most cases the first failure would have been in the next clock tick
2816 after the received time, and all would be well.
2818 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2819 being recognized when the domain list was tested by the match_domain
2820 condition in an expansion string.
2822 16. Added the ${str2b64: operator.
2824 17. Exim was always calling setrlimit() to set a large limit for the number of
2825 processes, without checking whether the existing limit was already
2826 adequate. (It did check for the limit on file descriptors.) Furthermore,
2827 errors from getrlimit() and setrlimit() were being ignored. Now they are
2828 logged to the main and panic logs, but Exim does carry on, to try to do its
2829 job under whatever limits there are.
2831 18. Imported PCRE 5.0.
2833 19. Trivial typo in log message " temporarily refused connection" (the leading
2836 20. If the log selector return_path_on_delivery was set and an address was
2837 redirected to /dev/null, the delivery process crashed because it assumed
2838 that a return path would always be set for a "successful" delivery. In this
2839 case, the whole delivery is bypassed as an optimization, and therefore no
2842 21. Internal re-arrangement: the function for sending a challenge and reading
2843 a response while authentication was assuming a zero-terminated challenge
2844 string. It's now changed to take a pointer and a length, to allow for
2845 binary data in such strings.
2847 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2849 23. Exim was not respecting finduser_retries when seeking the login of the
2850 uid under which it was called; it was always trying 10 times. (The default
2851 setting of finduser_retries is zero.) Also, it was sleeping after the final
2852 failure, which is pointless.
2854 24. Implemented tls_on_connect_ports.
2856 25. Implemented acl_smtp_predata.
2858 26. If the domain in control=submission is set empty, Exim assumes that the
2859 authenticated id is a complete email address when it generates From: or
2860 Sender: header lines.
2862 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2863 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2864 chown and chgrp in /bin and hostname in /usr/bin.
2866 28. Exim was keeping the "process log" file open after each use, just as it
2867 does for the main log. This opens the possibility of it remaining open for
2868 long periods when the USR1 signal hits a daemon. Occasional processlog
2869 errors were reported, that could have been caused by this. Anyway, it seems
2870 much more sensible not to leave this file open at all, so that is what now
2873 29. The long-running daemon process does not normally write to the log once it
2874 has entered its main loop, and it closes the log before doing so. This is
2875 so that log files can straightforwardly be renamed and moved. However,
2876 there are a couple of unusual error situations where the daemon does write
2877 log entries, and I had neglected to close the log afterwards.
2879 30. The text of an SMTP error response that was received during a remote
2880 delivery was being truncated at 512 bytes. This is too short for some of
2881 the long messages that one sometimes sees. I've increased the limit to
2884 31. It is now possible to make retry rules that apply only when a message has a
2885 specific sender, in particular, an empty sender.
2887 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2888 it possible to be selective about when SMTP synchronization is enforced.
2890 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2892 32. Implemented hosts_connection_nolog.
2894 33. Added an ACL for QUIT.
2896 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2899 35. Added mailbox_size and mailbox_filecount to appendfile.
2901 36. Added control = no_multiline_responses to ACLs.
2903 37. There was a bug in the logic of the code that waits for the clock to tick
2904 in the case where the clock went backwards by a substantial amount such
2905 that the microsecond fraction of "now" was more than the microsecond
2906 fraction of "then" (but the whole seconds number was less).
2908 38. Added support for the libradius Radius client library this is found on
2909 FreeBSD (previously only the radiusclient library was supported).
2915 1. When certain lookups returned multiple values in the form name=value, the
2916 quoting of the values was not always being done properly. Specifically:
2917 (a) If the value started with a double quote, but contained no whitespace,
2919 (b) If the value contained whitespace other than a space character (i.e.
2920 tabs or newlines or carriage returns) it was not quoted.
2921 This fix has been applied to the mysql and pgsql lookups by writing a
2922 separate quoting function and calling it from the lookup code. The fix
2923 should probably also be applied to nisplus, ibase and oracle lookups, but
2924 since I cannot test any of those, I have not disturbed their existing code.
2926 2. A hit in the callout cache for a specific address caused a log line with no
2927 reason for rejecting RCPT. Now it says "Previous (cached) callout
2928 verification failure".
2930 3. There was an off-by-one bug in the queryprogram router. An over-long
2931 return line was truncated at 256 instead of 255 characters, thereby
2932 overflowing its buffer with the terminating zero. As well as fixing this, I
2933 have increased the buffer size to 1024 (and made a note to document this).
2935 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2936 when Exim is waiting for an SMTP response from a remote server, Exim
2937 restarts its select() call on the socket, thereby resetting its timeout.
2938 This is not a problem when such interrupts are rare. Somebody set up a cron
2939 job to run exiwhat every 2 minutes, which is less than the normal select()
2940 timeout (5 or 10 minutes). This meant that the select() timeout never
2941 kicked in because it was always reset. I have fixed this by comparing the
2942 time when an interrupt arrives with the time at the start of the first call
2943 to select(). If more time than the timeout has elapsed, the interrupt is
2944 treated as a timeout.
2946 5. Some internal re-factoring in preparation for the addition of Sieve
2947 extensions (by MH). In particular, the "personal" test is moved to a
2948 separate function, and given an option for scanning Cc: and Bcc: (which is
2949 not set for Exim filters).
2951 6. When Exim created an email address using the login of the caller as the
2952 local part (e.g. when creating a From: or Sender: header line), it was not
2953 quoting the local part when it contained special characters such as @.
2955 7. Installed new OpenBSD configuration files.
2957 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2958 try to make them clearer.
2960 9. Callout options, other than the timeout value, were being ignored when
2961 verifying sender addresses in header lines. For example, when using
2963 verify = header_sender/callout=no_cache
2965 the cache was (incorrectly) being used.
2967 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2968 only the Cygwin environment.
2970 11. When return_path_on_delivery was set as a log selector, if different remote
2971 addresses in the same message used different return paths and parallel
2972 remote delivery occurred, the wrong values would sometimes be logged.
2973 (Whenever a remote delivery process finished, the return path value from
2974 the most recently started remote delivery process was logged.)
2976 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2977 lines when AUTH and/or TLS are in use. This is the "received protocol"
2978 field. Exim used to use "asmtp" for authenticated SMTP, without any
2979 indication (in the protocol name) for TLS use. Now it follows the RFC and
2980 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2981 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2982 names appear in log lines as well as in Received: header lines.
2984 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2985 extensions, and comparison tests, and to fix some bugs.
2987 14. Changes to the "personal" filter test:
2989 (1) The test was buggy in that it was just doing the equivalent of
2990 "contains" tests on header lines. For example, if a user's address was
2991 anne@some.where, the "personal" test would incorrectly be true for
2993 To: susanne@some.where
2995 This test is now done by extracting each address from the header in turn,
2996 and checking the entire address. Other tests that are part of "personal"
2997 are now done using regular expressions (for example, to check local parts
2998 of addresses in From: header lines).
3000 (2) The list of non-personal local parts in From: addresses has been
3001 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3002 taken from the Sieve specification recommendations.
3004 (3) If the message contains any header line starting with "List-" it is
3005 treated as non-personal.
3007 (4) The test for "circular" in the Subject: header line has been removed
3008 because it now seems ill-conceived.
3010 15. Minor typos in src/EDITME comments corrected.
3012 16. Installed latest exipick from John Jetmore.
3014 17. If headers_add on a router specified a text string that was too long for
3015 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3016 of string_sprintf() is now avoided.
3018 18. $message_body_size was not set (it was always zero) when running the DATA
3019 ACL and the local_scan() function.
3021 19. For the "mail" command in an Exim filter, no default was being set for
3022 the once_repeat time, causing a random time value to be used if "once" was
3023 specified. (If the value happened to be <= 0, no repeat happened.) The
3024 default is now 0s, meaning "never repeat". The "vacation" command was OK
3025 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3026 (I found it when inspecting the code).
3028 20. There is now an overall timeout for performing a callout verification. It
3029 defaults to 4 times the callout timeout, which applies to individual SMTP
3030 commands during the callout. The overall timeout applies when there is more
3031 than one host that can be tried. The timeout is checked before trying the
3032 next host. This prevents very long delays if there are a large number of
3033 hosts and all are timing out (e.g. when the network connections are timing
3034 out). The value of the overall timeout can be changed by specifying an
3035 additional sub-option for "callout", called "maxwait". For example:
3037 verify = sender/callout=5s,maxwait=20s
3039 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3040 to the end before writing, but this should make it even safer).
3042 22. Exim was forgetting that it had advertised PIPELINING for the second and
3043 subsequent messages on an SMTP connection. It was also not resetting its
3044 memory on STARTTLS and an internal HELO.
3046 23. When Exim logs an SMTP synchronization error within a session, it now
3047 records whether PIPELINING has been advertised or not.
3049 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3050 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3051 rather than long int.
3053 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3055 26. Added the never_mail option to autoreply.
3061 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3062 crash if the getsockname() call failed; this can happen if a connection is
3063 closed very soon after it is established. The problem was simply in the
3064 order in which certain operations were done, causing Exim to try to write
3065 to the SMTP stream before it had set up the file descriptor. The bug has
3066 been fixed by making things happen in the correct order.
3072 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3073 before the connection was closed, thus losing the rejection response.
3075 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3076 some early Solaris releases, but causes trouble in current releases where
3077 socklen_t is defined.
3079 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3082 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3085 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3086 that releasing the top of it at the end releases what was used for sub-
3087 expansions (unless the block got too big). However, discard this block if
3088 the first thing is a variable or header, so that we can use its block when
3089 it is dynamic (useful for very large $message_headers, for example).
3091 6. Lookups now cache *every* query, not just the most recent. A new, separate
3092 store pool is used for this. It can be recovered when all lookup caches are
3093 flushed. Lookups now release memory at the end of their result strings.
3094 This has involved some general refactoring of the lookup sources.
3096 7. Some code has been added to the store_xxx() functions to reduce the amount
3097 of flapping under certain conditions.
3099 8. log_incoming_interface used to affect only the <= reception log lines. Now
3100 it causes the local interface and port to be added to several more SMTP log
3101 lines, for example "SMTP connection from", and rejection lines.
3103 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3105 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3107 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3108 could be overwritten at the end of the current message (or the start of a
3109 new message if it was set in a HELO ACL). The value is now preserved for
3110 the duration of the SMTP connection.
3112 12. If a transport had a headers_rewrite setting, and a matching header line
3113 contained an unqualified address, that address was qualified, even if it
3114 did not match any rewriting rules. The underlying bug was that the values
3115 of the flags that permit the existence of unqualified sender and recipient
3116 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3117 non-local messages, and by -bnq for local messages) were not being
3118 preserved with the message after it was received.
3120 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3121 "next input=" as part of the text comprising the host identity instead of
3122 the correct text. The code was using the same buffer for two different
3123 strings. However, depending on which order the printing function evaluated
3124 its arguments, the bug did not always show up. Under Linux, for example, my
3125 test suite worked just fine.
3127 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3128 doesn't work with some older versions of Perl. It has been changed to "my",
3129 which in any case is probably the better facility to use.
3131 15. A really picky compiler found some instances of statements for creating
3132 error messages that either had too many or two few arguments for the format
3135 16. The size of the buffer for calls to the DNS resolver has been increased
3136 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3137 for addresses that have a lot of PTR records. This alleviates a problem; it
3138 does not fully solve it.
3140 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3141 buffer now truncates the list and logs the incident, which is the same
3142 action as happens when Exim is looking up a host name and its aliases.
3143 Previously in this situation something unpredictable would happen;
3144 sometimes it was "internal error: store_reset failed".
3146 18. If a server dropped the connection unexpectedly when an Exim client was
3147 using GnuTLS and trying to read a response, the client delivery process
3148 crashed while trying to generate an error log message.
3150 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3151 string, for example:
3153 warn message = H1: something\nH2: something
3155 the text was added as a single header line from Exim's point of view
3156 though it ended up OK in the delivered message. However, searching for the
3157 second and subsequent header lines using $h_h2: did not work. This has been
3158 fixed. Similarly, if a system filter added multiple headers in this way,
3159 the routers could not see them.
3161 20. Expanded the error message when iplsearch is called with an invalid key to
3162 suggest using net-iplsearch in a host list.
3164 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3165 ACLs are no longer actually imposed (and a message to that effect is
3168 22. If a "gecos" field in a passwd entry contained escaped characters, in
3169 particular, if it contained a \" sequence, Exim got it wrong when building
3170 a From: or a Sender: header from that name. A second bug also caused
3171 incorrect handling when an unquoted " was present following a character
3172 that needed quoting.
3174 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3175 was not being matched caselessly.
3177 24. Arranged for all hyphens in the exim.8 source to be escaped with
3180 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3181 properly. Recipient callout cache records were still being keyed to include
3182 the sender, even when use_sender was set false. This led to far more
3183 callouts that were necessary. The sender is no longer included in the key
3184 when use_sender is false.
3186 26. Added "control = submission" modifier to ACLs.
3188 27. Added the ${base62d: operator to decode base 62 numbers.
3190 28. dnsdb lookups can now access SRV records.
3192 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3193 the configuration file.
3195 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3196 (-v) mode. This makes the output for a verbose queue run more intelligible.
3198 31. Added a use_postmaster feature to recipient callouts.
3200 32. Added the $body_zerocount variable, containing the number of binary zero
3201 bytes in the message body.
3203 33. The time of last modification of the "new" subdirectory is now used as the
3204 "mailbox time last read" when there is a quota error for a maildir
3207 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3209 35. Added +ignore_unknown as a special item in host lists.
3211 36. Code for decoding IPv6 addresses in host lists is now included, even if
3212 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3213 address was recognized as an IP address, but was then not correctly decoded
3214 into binary, causing unexpected and incorrect effects when compared with
3221 1. Very minor rewording of debugging text in manualroute to say "list of
3222 hosts" instead of "hostlist".
3224 2. If verify=header_syntax was set, and a header line with an unqualified
3225 address (no domain) and a large number of spaces between the end of the
3226 name and the colon was received, the reception process suffered a buffer
3227 overflow, and (when I tested it) crashed. This was caused by some obsolete
3228 code that should have been removed. The fix is to remove it!
3230 3. When running in the test harness, delay a bit after writing a bounce
3231 message to get a bit more predictability in the log output.
3233 4. Added a call to search_tidyup() just before forking a reception process. In
3234 theory, someone could use a lookup in the expansion of smtp_accept_max_
3235 per_host which, without the tidyup, could leave open a database connection.
3237 5. Added the variables $recipient_data and $sender_data which get set from a
3238 lookup success in an ACL "recipients" or "senders" condition, or a router
3239 "senders" option, similar to $domain_data and $local_part_data.
3241 6. Moved the writing of debug_print from before to after the "senders" test
3244 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3245 problems for message scanning, either using a data ACL, or using
3246 local_scan() because the Received: header was not generated till after they
3247 were called (in order to set the time as the time of reception completion).
3248 I have revised the way this works. The header is now generated after the
3249 body is received, but before the ACL or local_scan() are called. After they
3250 are run, the timestamp in the header is updated.
3256 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3257 before starting a queue runner without re-exec. This happened only when
3258 deliver_drop_privilege was set or when the Exim user was set to root. The
3259 effect of the bug was that timeouts during subsequent deliveries caused
3260 crashes instead of being properly handled. The handler is now left at its
3261 default (and expected) setting.
3263 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3264 message, again when deliver_drop_privilege is set or Exim is run as root.
3265 The bug described in (1) was not present in this case, but the tidying up
3266 of the other signals was missing. I have made the two cases consistent.
3268 3. The ignore_target_hosts setting on a manualroute router was being ignored
3269 for hosts that were looked up using the /MX notation.
3271 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3274 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3275 operated on the sender address. After changing the $sender_address to <>
3276 for the sender address verify, Exim was re-instated it as the original
3277 (before rewriting) address, but remembering that it had rewritten it, so it
3278 wasn't rewriting it again. This bug also had the effect of breaking the
3279 sender address verification caching when the sender address was rewritten.
3281 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3282 This has been changed so that if the ip literal address matches
3283 ignore_target_hosts, the router declines.
3285 7. Added expansion conditions match_domain, match_address, and match_local_
3286 part (NOT match_host).
3288 8. The placeholder for the Received: header didn't have a length field set.
3290 9. Added code to Exim itself and to exim_lock to test for a specific race
3291 condition that could lead to file corruption when using MBX delivery. The
3292 issue is with the lockfile that is created in /tmp. If this file is removed
3293 after a process has opened it but before that process has acquired a lock,
3294 there is the potential for a second process to recreate the file and also
3295 acquire a lock. This could lead to two Exim processes writing to the file
3296 at the same time. The added code performs the same test as UW imapd; it
3297 checks after acquiring the lock that its file descriptor still refers to
3298 the same named file.
3300 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3301 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3304 11. Added the smtp_active_hostname option. If used, this will typically be made
3305 to depend on the incoming interface address. Because $interface_address is
3306 not set up until the daemon has forked a reception process, error responses
3307 that can happen earlier (such as "too many connections") no longer contain
3310 12. If an expansion in a condition on a "warn" statement fails because a lookup
3311 defers, the "warn" statement is abandoned, and the next ACL statement is
3312 processed. Previously this caused the whole ACL to be aborted.
3314 13. Added the iplsearch lookup type.
3316 14. Added ident_timeout as a log selector.
3318 15. Added tls_certificate_verified as a log selector.
3320 16. Added a global option tls_require_ciphers (compare the smtp transport
3321 option of the same name). This controls incoming TLS connections.
3323 17. I finally figured out how to make tls_require_ciphers do a similar thing
3324 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3325 before starting the TLS session.
3327 18. Tabs are now shown as \t in -bP output.
3329 19. If the log selector return_path_on_delivery was set, Exim crashed when
3330 bouncing a message because it had too many Received: header lines.
3332 20. If two routers both had headers_remove settings, and the first one included
3333 a superfluous trailing colon, the final name in the first list and the
3334 first name in the second list were incorrectly joined into one item (with a
3335 colon in the middle).
3341 1. Added -C and -D options to the exinext utility, mainly to make it easier
3342 to include in the automated testing, but these could be helpful when
3343 multiple configurations are in use.
3345 2. The exinext utility was not formatting the output nicely when there was
3346 an alternate port involved in the retry record key, nor when there was a
3347 message id as well (for retries that were specific to a specific message
3348 and a specific host). It was also confused by IPv6 addresses, because of
3349 the additional colons they contain. I have fixed the IPv4 problem, and
3350 patched it up to do a reasonable job for IPv6.
3352 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3353 delivery, the log line now contains "pipelined" if PIPELINING was used.
3355 4. An SMTP transport process used to panic and die if the bind() call to set
3356 an explicit outgoing interface failed. This has been changed; it is now
3357 treated in the same way as a connect() failure.
3359 5. A reference to $sender_host_name in the part of a conditional expansion
3360 that was being skipped was still causing a DNS lookup. This no longer
3363 6. The def: expansion condition was not recognizing references to header lines
3364 that used bh_ and bheader_.
3366 7. Added the _cache feature to named lists.
3368 8. The code for checking quota_filecount in the appendfile transport was
3369 allowing one more file than it should have been.
3371 9. For compatibility with Sendmail, the command line option
3379 and sets the incoming protocol and host name (for trusted callers). The
3380 host name and its colon can be omitted when only the protocol is to be set.
3381 Note the Exim already has two private options, -pd and -ps, that refer to
3382 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3383 "s", but I don't think that's a major issue.
3385 10. A number of refactoring changes to the code, none of which should affect
3388 (a) The number of logging options was getting close to filling up the
3389 32-bit word that was used as a bit map. I have split them into two classes:
3390 those that are passed in the argument to log_write(), and those that are
3391 only ever tested independently outside of that function. These are now in
3392 separate 32-bit words, so there is plenty of room for expansion again.
3393 There is no change in the user interface or the logging behaviour.
3395 (b) When building, for example, log lines, the code previously used a
3396 macro that called string_cat() twice, in order to add two strings. This is
3397 not really sufficiently general. Furthermore, there was one instance where
3398 it was actually wrong because one of the argument was used twice, and in
3399 one call a function was used. (As it happened, calling the function twice
3400 did not affect the overall behaviour.) The macro has been replaced by a
3401 function that can join an arbitrary number of extra strings onto a growing
3404 (c) The code for expansion conditions now uses a table and a binary chop
3405 instead of a serial search (which was left over from when there were very
3406 few conditions). Also, it now recognizes conditions like "pam" even when
3407 the relevant support is not compiled in: a suitably worded error message is
3408 given if an attempt is made to use such a condition.
3410 11. Added ${time_interval:xxxxx}.
3412 12. A bug was causing one of the ddress fields not to be passed back correctly
3413 from remote delivery subprocesses. The field in question was not being
3414 subsequently used, so this caused to problems in practice.
3416 13. Added new log selectors queue_time and deliver_time.
3418 14. Might have fixed a bug in maildirsizefile handling that threw up
3419 "unexpected character" debug warnings, and recalculated the data
3420 unnecessarily. In any case, I expanded the warning message to give more
3423 15. Added the message "Restricted characters in address" to the statements in
3424 the default ACL that block characters like @ and % in local parts.
3426 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3427 Three changes have been made:
3429 (a) There was a serious bug; a negative response to MAIL caused the whole
3430 recipient domain to be cached as invalid, thereby blocking all messages
3431 to all local parts at the same domain, from all senders. This bug has
3432 been fixed. The domain is no longer cached after a negative response to
3433 MAIL if the sender used is not empty.
3435 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3438 (c) A new callout option, "use_sender" has been added for people who want
3439 the modified behaviour.
3445 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3448 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3449 indeed breaks things for older releases.
3451 3. Added additional logging to the case where there is a problem reading data
3452 from a filter that is running in a subprocess using a pipe, in order to
3453 try to track down a specific problem.
3455 4. Testing facility fudge: when running in the test harness and attempting
3456 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3457 getting "No route to host". Convert this to a timeout.
3459 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3462 6. Some OS don't have socklen_t but use size_t instead. This affects the
3463 fifth argument of getsockopt() amongst other things. This is now
3464 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3465 can be set for individual OS. I have set it for SunOS5, OSF1, and
3466 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3467 some earlier ones do not.
3469 7. Change 4.30/15 was not doing the test caselessly.
3471 8. The standard form for an IPv6 address literal was being rejected by address
3472 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3473 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3474 this, as well as the form without the "IPv6" on the front (but only when
3475 address literals are enabled, of course).
3477 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3479 10. Exim crashed if a message with an empty sender address specified by -f
3480 encountered a router with an errors_to setting. This could be provoked only
3481 by a command such as
3485 where an empty string was supplied; "<>" did not hit this bug.
3487 11. Installed PCRE release 4.5.
3489 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3490 remained set. It is now erased.
3492 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3493 times from message ids (which are base 36 rather than the normal 62).
3495 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3496 were being counted as actual protocol errors, and logged if the log
3497 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3498 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3499 and DATA following a set of rejected RCPTs do not count as protocol errors.
3500 In other words, Exim assumes they were pipelined, though this may not
3501 actually be the case. Of course, in all cases the client gets an
3502 appropriate error code.
3504 15. If a lookup fails in an ACL condition, a message about the failure may
3505 be available; it is used if testing the ACL cannot continue, because most
3506 such messages specify what the cause of the deferral is. However, some
3507 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3508 that caused an old message to be retained and used if a later statement
3509 caused a defer, replacing the real cause of the deferral.
3511 16. If an IP address had so many PTR records that the DNS lookup buffer
3512 was not large enough to hold them, Exim could crash while trying to process
3513 the truncated data. It now detects and logs this case.
3515 17. Further to 4.21/58, another change has been made: if (and only if) the
3516 first line of a message (the first header line) ends with CRLF, a bare LF
3517 in a subsequent header line has a space inserted after it, so as not to
3518 terminate the header.
3520 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3521 unnecessarily, used atoi() instead of strtol(), and didn't check the
3522 termination when getting file sizes from file names by regex.
3524 19. Completely re-implemented the support for maildirsize files, in the light
3525 of a number of problems with the previous contributed implementation
3526 (4.30/29). In particular:
3528 . If the quota is zero, the maildirsize file is maintained, but no quota is
3531 . If the maildir directory does not exist, it is created before any attempt
3532 to write a maildirsize file.
3534 . The quota value in the file is just a cache; if the quota is changed in
3535 the transport, the new value overrides.
3537 . A regular expression is available for excluding directories from the
3540 20. The autoreply transport checks the characters in options that define the
3541 message's headers; it allows continued headers, but it was checking with
3542 isspace() after an embedded newline instead of explicitly looking for a
3545 21. If all the "regular" hosts to which an address was routed had passed their
3546 expiry times, and had not reached their retry times, the address was
3547 bounced, even if fallback hosts were defined. Now Exim should go on to try
3550 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3551 equivalent code in the SMTP transport. Some hosts send humungous responses
3552 to HELO/EHLO, more than 1024 it seems.
3554 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3555 gives compiler warnings in some environments. I've now done it "properly",
3558 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3559 (because of problems with the built-in one) was declared to return uschar *
3560 instead of char *, causing compiler failure.
3562 25. Fixed a file descriptor leak when processing alias/forward files.
3564 26. Fixed a minor format string issue in dbfn.c.
3566 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3568 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3569 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3572 29. When a sender address is verified, it is cached, to save repeating the test
3573 when there is more than one recipient in a message. However, when the
3574 verification involves a callout, it is possible for different callout
3575 options to be set for different recipients. It is too complicated to keep
3576 track of this in the cache, so now Exim always runs a verification when a
3577 callout is required, relying on the callout cache for the optimization.
3578 The overhead is duplication of the address routing, but this should not be
3581 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3582 to be verified with callout=postmaster, and the main callout worked but the
3583 postmaster check failed, the verification correctly failed. However, if a
3584 subsequent RCPT command asked for sender verification *without* the
3585 postmaster check, incorrect caching caused this verification also to fail,
3588 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3589 it was not caching the DNS options (qualify_single, search_parents) that
3590 were used when the lookup failed. A subsequent lookup with different
3591 options therefore always gave the same answer, though there were cases
3592 where it should not have. (Example: a "domains = !$mx_any" option on a
3593 dnslookup router: the "domains" option is always processed without any
3594 widening, but the router might have qualify_single set.) Now Exim uses the
3595 cached value only when the same options are set.
3597 32. Added John Jetmore's "exipick" utility to the distribution.
3599 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3600 than a timeout (e.g. a certificate is required, and is not provided), an
3601 Exim server now closes the connection immediately. Previously it waited for
3602 the client to close - but if the client is SSL, it seems that they each
3603 wait for each other, leading to a delay before one of them times out.
3605 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3606 maintained 0.8.x compatibility because I don't think many are using it, and
3607 it is clearly obsolete.
3609 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3612 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3613 client certificate was expired. A simple patch fixes this, though I don't
3614 understand the full logic of why the verify callback is called multiple
3617 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3618 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3619 which causes problems with some clients (such as the Certicom SSL Plus
3620 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3621 disables the coutermeasure allowing Eudora to connect."
3623 38. Exim was not checking that a write() to a log file succeeded. This could
3624 lead to Bad Things if a log got too big, in particular if it hit a file
3625 size limit. Exim now panics and dies if it cannot write to a log file, just
3626 as it does if it cannot open a log file.
3628 39. Modified OS/Makefile-Linux so that it now contains
3630 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3632 The two -D definitions ensure that Exim is compiled with large file
3633 support, which makes it possible to handle log files that are bigger than
3636 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3637 instance) a domain was checked against a named list that involved a lookup,
3638 causing $domain_data to be set, then another domain was checked against the
3639 same list, then the first domain was re-checked, the value of $domain_data
3640 after the final check could be wrong. In particular, if the second check
3641 failed, it could be set empty. This bug probably also applied to
3644 41. The strip_trailing_dot option was not being applied to the address given
3645 with the -f command-line option.
3647 42. The code for reading a message's header from the spool was incrementing
3648 $received_count, but never initializing it. This meant that the value was
3649 incorrect (doubled) while delivering a message in the same process in which
3650 it was received. In the most common configuration of Exim, this never
3651 happens - a fresh exec is done - but it can happen when
3652 deliver_drop_privilege is set.
3654 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3655 it now includes up to 150 characters of the unexpected data in the log
3658 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3659 and building data strings. The size of both of these buffers was 10 000
3660 bytes - far larger than anybody would *ever* want, thought I. Needless to
3661 say, somebody hit the limit. I have increased the maximum line length to
3662 20 000 and the maximum data length of concatenated lines to 100 000. I have
3663 also fixed two bugs, because there was no checking on these buffers. Tsk,
3664 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3665 buffer is too small.
3667 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3668 lsearch lookups. Now it does.
3670 46. When parsing a route_list item in a manualroute router, a fixed-length
3671 buffer was used for the list of hosts. I made this 1024 bytes long,
3672 thinking that nobody would ever have a list of hosts that long. Wrong.
3673 Somebody had a whole pile of complicated expansion conditions, and the
3674 string was silently truncated, leading to an expansion error. It turns out
3675 that it is easier to change to an unlimited length (owing to other changes
3676 that have happened since this code was originally written) than to build
3677 structure for giving a limitation error. The length of the item that
3678 expands into the list of hosts is now unlimited.
3680 47. The lsearch lookup could not handle data where the length of text line was
3681 more than 4095 characters. Such lines were truncated, leading to shortened
3682 data being returned. It should now handle lines of any length.
3684 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3685 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3688 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3689 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3690 no longer happens in comment lines. A list of replacements is now placed
3691 at the head of all of the source files, except those whose only change is
3692 to replace PERL_COMMAND in the very first #! line.
3694 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3695 messages on the queue, with a bottom-up merge sort, using code contributed
3696 by Michael Haardt. This should make operations like -bp somewhat faster on
3697 large queues. It won't affect queue runners, except when queue_run_in_order
3700 51. Installed eximstats 1.31 in the distribution.
3702 52. Added support for SRV lookups to the dnslookup router.
3704 53. If an ACL referred to $message_body or $message_body_end, the value was not
3705 reset for any messages that followed in the same SMTP session.
3707 54. The store-handling optimization for building very long strings was not
3708 differentiating between the different store pools. I don't think this
3709 actually made any difference in practice, but I've tidied it.
3711 55. While running the routers to verify a sender address, $sender_address
3712 was still set to the sender address. This is wrong, because when routing to
3713 send a bounce to the sender, it would be empty. Therefore, I have changed
3714 it so that, while verifying a sender address, $sender_address is set to <>.
3715 (There is no change to what happens when verifying a recipient address.)
3717 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3718 target A or AAAA records (if not already returned) without resetting the
3719 qualify_single or search_parents options of the DNS resolver. These are
3720 inappropriate in this case because the targets of MX and SRV records must
3721 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3722 target that, when qualified, matched something in the local domain. These
3723 two options are now turned off when doing these lookups.
3725 57. It seems that at least some releases of Reiserfs (which does not have the
3726 concept of a fixed number of inodes) returns zero and not -1 for the
3727 number of available inodes. This interacted badly with check_spool_inodes,
3728 which assumed that -1 was the "no such thing" setting. What I have done is
3729 to check that the total number of inodes is greater than zero before doing
3730 the test of how many are available.
3732 58. When a "warn" ACL statement has a log_message modifier, the message is
3733 remembered, and not repeated. This is to avoid a lot of repetition when a
3734 message has many recipients that cause the same warning to be written.
3735 Howewer, Exim was preserving the list of already written lines for an
3736 entire SMTP session, which doesn't seem right. The memory is now reset if a
3737 new message is started.
3739 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3740 debugging output unless log_rewrite was also set.
3742 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3743 of (int)(handle) when we know that handle contains (void *)(-1).
3745 61. The Exim daemon panic-logs an error return when it closes the incoming
3746 connection. However "connection reset by peer" seems to be common, and
3747 isn't really an error worthy of noting specially, so that particular error
3750 62. When Exim is trying to find all the local interfaces, it used to panic and
3751 die if the ioctl to get the interface flags failed. However, it seems that
3752 on at least one OS (Solaris 9) it is possible to have an interface that is
3753 included in the list of interfaces, but for which you get a failure error
3754 for this call. This happens when the interface is not "plumbed" into a
3755 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3756 failure of the "get flags" call assumes that the interface is down.
3758 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3759 makes life easier for people who are doing arithmetic on fields extracted
3760 from dates, where you often get leading zeros that should not be
3761 interpreted as octal.
3763 64. Added qualify_domain to the redirect router, to override the global
3766 65. If a pathologically long header line contained very many addresses (the
3767 report of this problem mentioned 10 000) and each of them was rewritten,
3768 Exim could use up a very large amount of memory. (It kept on making new
3769 copies of the header line as it rewrote, and never released the old ones.)
3770 At the expense of a bit more processing, the header rewriting function has
3771 been changed so that it no longer eats memory in this way.
3773 66. The generation of the Received: header has been moved from the time that a
3774 message starts to be received, to the time that it finishes. The timestamp
3775 in the Received: header should now be very close to that of the <= log
3776 line. There are two side-effects of this change:
3778 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3779 logged header lines no longer include the local Received: line, because
3780 it has not yet been created. The same applies to a copy of the message
3781 that is returned to a non-SMTP sender when a message is rejected.
3783 (b) When a filter file is tested using -bf, no additional Received: header
3784 is added to the test message. After some thought, I decided that this
3787 This change does not affect the value of $received_for. It is still set
3788 after address rewriting, but before local_scan() is called.
3790 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3792 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3793 gave an unhelpful panic error message, and a defer error. I have managed to
3794 change this behaviour so that it now rejects any supplied certificate,
3795 which seems right, as the list of acceptable certificates is empty.
3797 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3798 gave an unhelpful defer error. I have not managed to make this reject any
3799 supplied certificates, but the error message it gives is "no certificate
3800 supplied", which is not helpful.
3802 70. exigrep's output now also includes lines that are not associated with any
3803 message, but which match the given pattern. Implemented by a patch from
3804 Martin Sluka, which also tidied up the Perl a bit.
3806 71. Recipient callout verification, like sender verification, was using <> in
3807 the MAIL FROM command. This isn't really the right thing, since the actual
3808 sender may affect whether the remote host accepts the recipient or not. I
3809 have changed it to use the actual sender in the callout; this means that
3810 the cache record is now keyed on a recipient/sender pair, not just the
3811 recipient address. There doesn't seem to be a real danger of callout loops,
3812 since a callout by the remote host to check the sender would use <>.
3813 [SEE ABOVE: changed after hitting problems.]
3815 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3816 temporary errors. However, in the case of such a code being given after
3817 the end of a data transmission (i.e. after ".") Exim was failing to write
3818 a retry record for the message. (Yes, there was some broken host that was
3819 actually sending 8xx at this point.)
3821 73. An unknown lookup type in a host list could cause Exim to panic-die when
3822 the list was checked. (An example that provoked this was putting <; in the
3823 middle of a list instead of at the start.) If this happened during a DATA
3824 ACL check, a -D file could be left lying around. This kind of configuration
3825 error no longer causes Exim to die; instead it causes a defer errror. The
3826 incident is still logged to the main and panic logs.
3828 74. Buglet left over from Exim 3 conversion. The message "too many messages
3829 in one connection" was written to the rejectlog but not the mainlog, except
3830 when address rewriting (yes!) was being logged.
3832 75. Added write_rejectlog option.
3834 76. When a system filter was run not as root (that is, when system_filter_user
3835 was set), the values of the $n variables were not being returned to the
3836 main process; thus, they were not subsequently available in the $sn
3839 77. Added +return_path_on_delivery log selector.
3841 78. A connection timeout was being treated differently from recipients deferred
3842 when testing hosts_max_try with a message that was older than the host's
3843 retry timeout. (The host should not be counted, thus allowing all hosts to
3844 be tried at least once before bouncing.) This may have been the cause of an
3845 occasionally reported bug whereby a message would remain on the queue
3846 longer than the retry timeout, but would be bounced if a delivery was
3847 forced. I say "may" because I never totally pinned down the problem;
3848 setting up timeout/retry tests is difficult. See also the next item.
3850 79. The ultimate address timeout was not being applied to errors that involved
3851 a combination of host plus message (for example, a timeout on a MAIL
3852 command). When an address resolved to a number of possible hosts, and they
3853 were not all tried for each delivery (e.g. because of hosts_max_try), a
3854 message could remain on the queue longer than the retry timeout.
3856 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3859 81. Fixed an obscure SMTP outgoing bug which required at least the following
3860 conditions: (a) there was another message waiting for the same server;
3861 (b) the server returned 5xx to all RCPT commands in the first message so
3862 that the message was not completed; (c) the server dropped the connection
3863 or gave a negative response to the RSET that Exim sends to abort the
3864 transaction. The observed case was a dropped connection after DATA that had
3865 been sent in pipelining mode. That is, the server had advertised PIPELINING
3866 but was not implementing it correctly. The effect of the bug was incorrect
3867 behaviour, such as trying another host, and this could lead to a crash.
3873 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3874 and daemon.c were passed as pointers to ints; they should have been
3875 pointers to socklen_t variables (which are typically unsigned ints).
3877 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3880 3. Fixed a really odd bug that affected only the testing scheme; patching a
3881 certain fixed string in the binary changed the value of another string that
3882 happened to be identical to the end of the original first string.
3884 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3885 name", it returns that address as the IP address. On some operating
3886 systems (e.g. Solaris), it also passes back the IP address string as the
3887 "host name". However, on others (e.g. Linux), it passes back an empty
3888 string. Exim wasn't checking for this, and was changing the host name to an
3889 empty string, assuming it had been canonicized.
3891 5. Although rare, it is permitted to have more than one PTR record for a given
3892 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3893 all the names associated with an address, because they do in Solaris.
3894 However, it seems that they do not in Linux for data that comes from the
3895 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3896 I found this out when I moved to a new Linux workstation and tried to run
3897 the Exim test suite.
3899 To get round this problem I have changed the code so that it now does its
3900 own call to the DNS to look up PTR records when searching for a host name.
3901 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3902 addresses that are only in /etc/hosts are still found.
3904 This behaviour is, however, controlled by an option called host_lookup_
3905 order, which defaults to "bydns:byaddr". If people want to use the other
3906 order, or indeed, just use one or the other means of lookup, they can
3907 specify it in this variable.
3909 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3910 some operating systems, this comes back from gethostbyaddr() as an empty
3911 string, and this is what Exim used to test for. However, it seems that in
3912 other systems, "." is yielded. Exim now tests for this case too.
3914 7. The values of check_spool_space and check_log_space are now held internally
3915 as a number of kilobytes instead of an absolute number of bytes. If a
3916 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3917 kilobyte. This means that much larger values can be stored.
3919 8. Exim monitor: an attempt to get the action menu when not actually pointing
3920 at a message produces an empty menu entitled "No message selected". This
3921 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3922 no entries in it ("Shell widget menu has zero width and/or height"). So I
3923 have added a single, blank menu entry in this case.
3925 9. Added ${quote_local_part.
3927 10. MIME decoding is now applied to the contents of Subject: header lines when
3930 11. Now that a reference to $sender_host_address automatically causes a reverse
3931 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3932 host lookup before query-style lookups in lists that might use this
3933 variable. This has therefore been abolished, and the "net-" prefix is no
3934 longer necessary for query-style lookups.
3936 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3937 have been a typo for LFLAGS, so it has been changed.
3939 13. The install script calls Exim with "-C /dev/null" in order to find the
3940 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3941 to be output. Howeve, since Exim outputs its version number before the
3942 error, it didn't break the script. It just looked ugly. I fixed this by
3943 always allowing "-C /dev/null" if the caller is root.
3945 14. Ignore overlarge ACL variable number when reading spool file - insurance
3946 against a later release with more variables having written the file.
3948 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3949 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3950 this, as well as the form without the "IPv6" on the front.
3952 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3953 OS/Makefile-Darwin file.
3955 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3956 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3958 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3959 variable. However, some OS use ints for inodes. I've added cast to long int
3960 to get rid of the compiler warning.
3962 19. I had forgotten to lock out "/../" in configuration file names when
3963 ALT_CONFIG_PREFIX was set.
3965 20. Routers used for verification do not need to specify transports. However,
3966 if such a router generated a host list, and callout was configured, Exim
3967 crashed, because it could not find a port number from the (non-existent)
3968 transport. It now assumes port 25 in this circumstance.
3970 21. Added the -t option to exigrep.
3972 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3973 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3974 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3975 them is set, LOOKUP_LSEARCH is forced.
3977 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3979 24. Added sender and host information to the "rejected by local_scan()" log
3980 line; previously there was no indication of these.
3982 25. Added .include_if_exists.
3984 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3985 Linux. It turns out that not all file systems support this. Apparently some
3986 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3987 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3988 not supported on the file descriptor, is now ignored when Exim is trying to
3989 sync a directory. This applies only to Linux.
3991 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3993 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3994 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3997 29. Added support for maildirsize files from supplied patch (modified a bit).
3999 30. The use of :fail: followed by an empty string could lead Exim to respond to
4000 sender verification failures with (e.g.):
4002 550 Verification failed for <xxx>
4003 550 Sender verify failed
4005 where the first response line was missing the '-' that indicates it is not
4006 the final line of the response.
4008 31. The loop for finding the name of the user that called Exim had a hardwired
4009 limit of 10; it now uses the value of finduser_retries, which is used for
4010 all other user lookups.
4012 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4015 33. Exim was neglecting to zero errno before one call of strtol() when
4016 expanding a string and expecting an integer value. On some systems this
4017 resulted in spurious "integer overflow" errors. Also, it was casting the
4018 result into an int without checking.
4020 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4021 did not work. The code looks as if it has *never* worked, though it appears
4022 to have been documented since at least releast 1.62. I have made it work.
4024 35. The "timeout_DNS" error in retry rules, also documented since at least
4025 1.62, also never worked. As it isn't clear exactly what this means, and
4026 clearly it isn't a major issue, I have abolished the feature by treating it
4027 as "timeout", and writing a warning to the main and panic logs.
4029 36. The display of retry rules for -brt wasn't always showing the error code
4032 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4033 timeout_connect_A, timeout_connect_MX.
4035 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4036 to the empty sender.
4038 39. The daemon was not analysing the content of -oX till after it had closed
4039 stderr and disconnected from the controlling terminal. This meant that any
4040 syntax errors were only noted on the panic log, and the return code from
4041 the command was 0. By re-arranging the code a little, I've made the
4042 decoding happen first, so such errors now appear on stderr, and the return
4043 code is 1. However, the actual setting up of the sockets still happens in
4044 the disconnected process, so errors there are still only recorded on the
4047 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4048 connections (as happens on some IP stacks) was logged at start up time as
4049 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4050 it from "IPv6 and IPv4", which means that two separate sockets are being
4053 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4054 says whether AF_INET or AF_INET6 was passed as an argument.
4056 42. Exiwhat output was messed up when time zones were included in log
4059 43. Exiwhat now gives more information about the daemon's listening ports,
4060 and whether -tls-on-connect was used.
4062 44. The "port" option of the smtp transport is now expanded.
4064 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4065 silently ignored. Now an error message is written to the main and panic
4068 46. There's a new ACL modifier called "logwrite" which writes to a log file
4069 as soon as it is encountered.
4071 47. Added $local_user_uid and $local_user_gid at routing time.
4073 48. Exim crashed when trying to verify a sender address that was being
4076 49. Exim was recognizing only a space character after ".include". It now also
4077 recognizes a tab character.
4079 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4080 extracting the relevant information from the specification. The man page no
4081 longer contains scrambled data for the -d option, and I've added a section
4082 at the front about calling Exim under different names.
4084 51. Added "extra_headers" argument to the "mail" command in filter files.
4086 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4089 53. Installed eximstats 1.29.
4091 54. Added transport_filter_timeout as a generic transport option.
4093 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4094 Cc: header lines. This was required by RFC 822, but it not required by RFC
4097 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4098 incoming messages that did not have them. Now it does so only if the
4099 message originates locally, that is, if there is no associated remote host
4100 address. When Resent- header lines are present, this applies to the Resent-
4101 lines rather than the non-Resent- lines.
4103 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4104 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4106 58. Messages for configuration errors now include the name of the main
4107 configuration file - useful now that there may be more than one file in a
4108 list (.included file names were always shown).
4110 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4111 for those rare installations that do not start the daemon as root or run it
4112 setuid root. I've cut out the call to initgroups() if the daemon is not
4115 60. The Exim user and group can now be bound into the binary as text strings
4116 that are looked up at the start of Exim's processing.
4118 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4120 62. Added $mailstore_basename variable.
4122 63. Installed patch to sieve.c from Michael Haardt.
4124 64. When Exim failed to open the panic log after failing to open the main log,
4125 the original message it was trying to log was written to stderr and debug
4126 output, but if they were not available (the usual case in production), it
4127 was lost. Now it is written to syslog before the two lines that record the
4128 failures to open the logs.
4130 65. Users' Exim filters run in subprocesses under the user's uid. It is
4131 possible for a "deliver" command or an alias in a "personal" command to
4132 provoke an address rewrite. If logging of address rewriting is configured,
4133 this fails because the process is not running as root or exim. There may be
4134 a better way of dealing with this, but for the moment (because 4.30 needs
4135 to be released), I have disabled address rewrite logging when running a
4136 filter in a non-root, non-exim process.
4142 1. The buildconfig auxiliary program wasn't quoting the value set for
4143 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4144 not defined. This bug was masked in 4.22 by the effect that was fixed in
4147 2. Some messages that were rejected after a message id was allocated were
4148 shown as "incomplete" by exigrep. It no longer does this for messages that
4149 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4151 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4152 have allow_domain_literals set, the ID did not get logged in the <= line.
4153 Domain literals are now always recognized in Message-ID: header lines.
4155 4. The first argument for a ${extract expansion item is the key name or field
4156 number. Leading and trailing spaces in this item were not being ignored,
4157 causing some misleading effects.
4159 5. When deliver_drop_privilege was set, single queue runner processes started
4160 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4161 same command in the process it spins off) were not dropping privilege.
4163 6. When the daemon running as "exim" started a queue runner, it always
4164 re-executed Exim in the spun-off process. This is a waste of effort when
4165 deliver_drop_privilege is set. The new process now just calls the
4166 queue-runner function directly.
4172 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4175 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4176 ignored. Though the use of -oP was forcing the writing of a pid file, it
4177 was always written to the default place.
4179 3. If the message "no IP address found for host xxxx" is generated during
4180 incoming verification, it is now followed by identification of the incoming
4181 connection (so you can more easily find what provoked it).
4183 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4185 5. Added some features to "harden" Exim a bit more against certain attacks:
4187 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4188 be put in Local/Makefile. This is like the never_users runtime option,
4189 but it cannot be overridden. The default setting is "root".
4191 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4192 prefix string with which any file named in a -C command line option
4195 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4196 is retained for -C and -D only if the caller of Exim is root. Without
4197 it, the exim user may also use -C and -D and retain privilege.
4199 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4200 command line option is disabled.
4202 6. Macro names set by the -D option must start with an upper case letter, just
4203 like macro names defined in the configuration file.
4205 7. Added "dereference=" facility to LDAP.
4207 8. Two instances of the typo "uknown" in the source files are fixed.
4209 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4210 the Configure-Makefile script screwed up while processing it.
4212 10. Incorporated PCRE 4.4.
4214 11. The SMTP synchronization check was not operating right at the start of an
4215 SMTP session. For example, it could not catch a HELO sent before the client
4216 waited for the greeting. There is now a check for outstanding input at the
4217 point when the greeting is written. Because of the duplex, asynchronous
4218 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4219 way, but not yet received, when the check is performed.
4221 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4222 on TCP/IP sockets, because this apparently causes some broken clients to
4225 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4226 unchanged) from the Cygwin maintainer.
4228 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4229 was supported instead of testing for mbx. Effectively a typo.
4231 15. The spa authenticator server code was not checking that the input it
4232 received was valid base64.
4234 16. The debug output line for the "set" modifier in ACLs was not showing the
4235 name of the variable that was being set.
4237 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4239 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4240 lookup on its own. Something else was needed to trigger the lookup. For
4241 example, a match in host_lookup or the need for a host name in a host list.
4242 Now, if $sender_host_name is referenced and the host name has not yet been
4243 looked up, a lookup is performed. If the lookup fails, the variable remains
4244 empty, and $host_lookup_failed is set to "1".
4246 19. Added "eqi" as a case-independent comparison operator.
4248 20. The saslauthd authentication condition could segfault if neither service
4249 nor realm was specified.
4251 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4252 error message that was logged was misleading, and incoming SMTP
4253 connections were dropped. The message is now more accurate, and temporary
4254 errors are given to SMTP connections.
4256 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4257 (except RSET) with a 421 error, until QUIT is received. However, it was
4258 failing to send a response to QUIT.
4260 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4261 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4262 ACL is now run after the helo_try_verify_hosts code.
4264 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4265 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4266 case-independent, so other case variants are also recognized). Apparently
4267 some systems use these upper case variants.
4269 25. If more than two messages were waiting for the same host, and a transport
4270 filter was specified for the transport, Exim sent two messages over the
4271 same TCP/IP connection, and then failed with "socket operation on non-
4272 socket" when it tried to send the third.
4274 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4276 27. The extern definition of crypt16() in expand.c was not being excluded when
4277 the OS had its own crypt16() function.
4279 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4280 as a preferred synonym for return_size_limit, both as an option and as an
4283 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4285 30. Changed the default configuration ACL to relax the local part checking rule
4286 for addresses that are not in any local domains. For these addresses,
4287 slashes and pipe symbols are allowed within local parts, but the sequence
4288 /../ is explicitly forbidden.
4290 31. SPA server authentication was not clearing the challenge buffer before
4293 32. log_message in a "warn" ACL statement was writing to the reject log as
4294 well as to the main log, which contradicts the documentation and doesn't
4295 seem right (because no rejection is happening). So I have stopped it.
4297 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4298 However, I am unable to do any testing of this.
4300 34. Fixed an infelicity in the appendfile transport. When checking directories
4301 for a mailbox, to see if any needed to be created, it was accidentally
4302 using path names with one or more superfluous leading slashes; tracing
4303 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4305 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4306 contents are added to the log line that is written for every discarded
4307 recipient. (Previously a log_message setting was ignored.)
4309 36. The ${quote: operator now quotes the string if it is empty.
4311 37. The install script runs exim in order to find its version number. If for
4312 some reason other than non-existence or emptiness, which it checks, it
4313 could not run './exim', it was installing it with an empty version number,
4314 i.e. as "exim-". This error state is now caught, and the installation is
4317 38. An argument was missing from the function that creates an error message
4318 when Exim fails to connect to the socket for saslauthd authentication.
4319 This could cause Exim to crash, or give a corrupted message.
4321 39. Added isip, isip4, and isip6 to ${if conditions.
4323 40. The ACL variables $acl_xx are now saved with the message, and can be
4324 accessed later in routers, transports, and filters.
4326 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4327 strings in the file are not string-expanded.
4329 42. If a MAIL command specified a SIZE value that was too large to fit into an
4330 int variable, the check against message_size_limit failed. Such values are
4331 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4332 day this will have to be increased, but I don't think I want to be around
4333 when emails are that large.
4340 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4341 iconv() is not standard in FreeBSD.
4343 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4344 IPv6 enabled. The observed symptom was a segmentation fault on return from
4345 the function os_common_find_running_interfaces() in src/os.c.
4347 3. In the check_special_case() function in daemon.c I had used "errno" as an
4348 argument name, which causes warnings on some systems. This was basically a
4349 typo, since it was named "eno" in the comments!
4351 4. The code that waits for the clock to tick (at a resolution of some fraction
4352 of a second) so as to ensure message-id uniqueness was always waiting for
4353 at least one whole tick, when it could have waited for less. [This is
4354 almost certainly not relevant at current processor speeds, where it is
4355 unlikely to ever wait at all. But we try to future-proof.]
4357 5. The function that sleeps for a time interval that includes fractions of a
4358 second contained a race. It did not block SIGALRM between setting the
4359 timer, and suspending (a couple of lines later). If the interval was short
4360 and the sigsuspend() was delayed until after it had expired, the suspension
4361 never ended. On busy systems this could lead to processes getting stuck for
4364 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4365 process, before it forks any delivery processes. The open lookup caching
4366 mechanism meant that the open file or database connection was passed into
4367 the delivery process. The problem was that delivery processes always tidy
4368 up cached lookup data. This could cause a problem for the next delivery
4369 process started by the queue runner, because the external queue runner
4370 process does not know about the closure. So the next delivery process
4371 still has data in the lookup cache. In the case of a file lookup, there was
4372 no problem because closing a file descriptor in a subprocess doesn't affect
4373 the parent. However, if the lookup was caching a connection to a database,
4374 the connection was closed, and the second delivery process was likely to
4375 see errors such as "PGSQL: query failed: server closed the connection
4376 unexpectedly". The problem has been fixed by closing all cached lookups
4377 in a queue runner before running a delivery process.
4379 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4380 seem to have the "const" qualifier which it has on other OS. I've
4383 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4384 *of the same type* (client or server) with the same public name that an
4385 error should be diagnosed.
4387 9. When Exim looked up a host name for an IP address, but failed to find the
4388 original IP address when looking up the host name (a safety check), it
4389 output the message "<ip address> does not match any IP for NULL", which was
4390 confusing, to say the least. The bug was that the host name should have
4391 appeared instead of "NULL".
4393 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4394 user that is built into the binary, and the -C or -D options is used, root
4395 privilege is dropped before the configuration file is read. In addition,
4396 logging is switched to stderr instead of the normal log files. If the
4397 configuration then re-defines the Exim user, the unprivileged environment
4398 is probably not what is expected, so Exim logs a panic warning message (but
4401 However, if deliver_drop_privilege is set, the unprivileged state may well
4402 be exactly what is intended, so the warning has been cut out in that case,
4403 and Exim is allowed to try to write to its normal log files.
4409 1. smtp_return_error_details was not giving details for temporary sender
4410 or receiver verification errors.
4412 2. Diagnose a configuration error if two authenticators have the same public
4415 3. Exim used not to create the message log file for a message until the first
4416 delivery attempt. This could be confusing when incoming messages were held
4417 for policy or load reasons. The message log file is now created at the time
4418 the message is received, and an initial "Received" line is written to it.
4420 4. The automatically generated man page for command line options had a minor
4421 bug that caused no ill effects; however, a more serious problem was that
4422 the procedure for building the man page automatically didn't always
4423 operate. Consequently, release 4.20 contains an out-of-date version. This
4424 shouldn't happen again.
4426 5. When building Exim with embedded Perl support, the script that builds the
4427 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4428 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4430 6. The freeze_tell option was not being used for messages that were frozen on
4431 arrival, either by an ACL or by local_scan().
4433 7. Added the smtp_incomplete_transaction log selector.
4435 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4436 was accepting AUTH without a new EHLO.
4438 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4439 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4442 10. Logging of TCP/IP connections (when configured) now happens in the main
4443 daemon process instead of the child process, so that the TCP/IP connection
4444 count is more accurate (but it can never be perfect).
4446 11. The use of "drop" in a nested ACL was not being handled correctly in the
4447 outer ACL. Now, if condition failure induced by the nested "drop" causes
4448 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4449 or "require"), the connection is dropped.
4451 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4452 that yield "discard" can now be used with an "accept" or a "discard" verb,
4453 but an error is generated for any others (because I can't see a useful way
4454 to define what should happen).
4456 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4457 are now processed in the same way as lines in the Exim configuration file.
4458 In particular, continuation lines are supported.
4460 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4462 15. Added -ti meaning -t -i.
4464 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4465 domains, and warn by logging if others are found.
4467 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4468 returned from GIFCONF when Exim is trying to find the list of interfaces on
4469 a host. The code in os.c has been modified to copy each ifreq to an aligned
4470 structure in all cases.
4472 Also, in some cases, the returned ifreq's were being copied to a 'struct
4473 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4474 means the last couple of bytes of an IPv6 address could be chopped if the
4475 ifreq contained only a normal sockaddr (14 bytes storage).
4477 18. Named domain lists were not supported in the hosts_treat_as_local option.
4478 An entry such as +xxxx was not recognized, and was treated as a literal
4481 19. Ensure that header lines added by a DATA ACL are included in the reject log
4482 if the ACL subsequently rejects the message.
4484 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4485 MD5 (which is deprecated).
4487 21. When testing a filter file using -bf, Exim was writing a message when it
4488 took the sender from a "From " line in the message, but it was not doing so
4489 when it took $return_path from a Return-Path: header line. It now does.
4491 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4492 with a valid header line field name (a series of printing characters
4493 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4495 23. Changed "disc" in the source to "disk" to conform to the documentation and
4496 the book and for uniformity.
4498 24. Ignore Sendmail's -Ooption=value command line item.
4500 25. When execve() failed while trying to run a command in a pipe transport,
4501 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4502 could be confused with a return value of 69 from the command itself. This
4503 has been changed to 127, the value the shell returns if it is asked to run
4504 a non-existent command. The wording for the related log line suggests a
4505 non-existent command as the problem.
4507 26. If received_header_text expands to an empty string, do not add a Received:
4508 header line to the message. (Well, it adds a token one on the spool, but
4509 marks it "old" so that it doesn't get used or transmitted.)
4511 27. Installed eximstats 1.28 (addition of -nt option).
4513 28. There was no check for failure on the call to getsockname() in the daemon
4514 code. This can fail if there is a shortage of resources on the system, with
4515 ENOMEM, for example. A temporary error is now given on failure.
4517 29. Contrary to the C standard, it seems that in some environments, the
4518 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4519 program. Exim now does this explicitly; it affects the formatting of
4520 timestamps using strftime().
4522 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4523 complaints. I've now initialized all the variables, to avoid this.
4525 32. Header lines added by a system filter were not being "seen" during
4526 transport-time rewrites.
4528 33. The info_callback() function passed to OpenSSL is set up with type void
4529 (*)(SSL *, int, int), as described somewhere. However, when calling the
4530 function (actually a macro) that sets it up, the type void(*)() is
4531 expected. I've put in a cast to prevent warnings from picky compilers.
4533 34. If a DNS black list lookup found a CNAME record, but there were no A
4534 records associated with the domain it pointed at, Exim crashed.
4536 35. If a DNS black list lookup returned more than one A record, Exim ignored
4537 all but the first. It now scans all returned addresses if a particular IP
4538 value is being sought. In this situation, the contents of the
4539 $dnslist_value variable are a list of all the addresses, separated by a
4542 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4543 to accept a host name and all its aliases if the forward lookup for any of
4544 them yielded the IP address of the incoming connection. Now it accepts only
4545 those names whose forward lookup yields the correct IP address. Any other
4546 names are discarded. This closes a loophole whereby a rogue DNS
4547 administrator could create reverse DNS records to break through a
4548 wildcarded host restriction in an ACL.
4550 37. If a user filter or a system filter that ran in a subprocess used any of
4551 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4552 the wrong values were passed to the pipe command ($thisaddress had the
4553 value of $0, $0 had the value of $1, etc). This bug was introduced by
4554 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4556 38. Improved the line breaking for long SMTP error messages from ACLs.
4557 Previously, if there was no break point between 40 and 75 characters, Exim
4558 left the rest of the message alone. Two changes have been made: (a) I've
4559 reduced the minimum length to 35 characters; (b) if it can't find a break
4560 point between 35 and 75 characters, it looks ahead and uses the first one
4561 that it finds. This may give the occasional overlong line, but at least the
4562 remaining text gets split now.
4564 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4565 file descriptors might be low, and that setting 1000 would always raise it.
4566 It turns out that in some environments, the limit is already over 1000 and
4567 that lowering it causes trouble. So now Exim takes care not to decrease it.
4569 40. When delivering a message, the value of $return_path is set to $sender_
4570 address at the start of routing (routers may change the value). By an
4571 oversight, this default was not being set up when an address was tested by
4572 -bt or -bv, which affected the outcome if any router or filter referred to
4575 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4576 log only when "message" or "log_message" are set. However, if one of the
4577 conditions was an address verification, or a call to a nested ACL, the
4578 messages generated by the underlying test were being passed through. This
4579 no longer happens. The underlying message is available in $acl_verify_
4580 message for both "message" and "log_message" expansions, so it can be
4581 passed through if needed.
4583 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4584 new expansion $bh_ to give the encoded byte string without charset
4585 translation. Translation happens only if iconv() is available; HAVE_ICONV
4586 indicates this at build time. HEADERS_CHARSET gives the charset to
4587 translate to; headers_charset can change it in the configuration, and
4588 "headers charset" can change it in an individual filter file.
4590 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4591 that creates RFC 2047 encoded "words" labels them as that charset instead
4592 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4593 expansion operator; (ii) when Exim creates a From: line for a local
4594 message; (iii) when a header line is rewritten to include a "phrase" part.
4596 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4597 buggy, causing it to skip the first lines of messages whose message ID
4598 ended in 'D'. This would not have bitten before Exim release 4.14, because
4599 message IDs were unlikely to end in 'D' before then. The effect was to have
4600 incorrect size information for certain domains.
4602 45. #include "config.h" was missing at the start of the crypt16.c module. This
4603 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4606 46. If there was a timeout during a "random" callout check, Exim treated it as
4607 a failure of the random address, and carried on sending RSET and the real
4608 address. If the delay was just some slowness somewhere, the response to the
4609 original RCPT would be taken as a response to RSET and so on, causing
4610 mayhem of various kinds.
4612 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4613 when I implemented it. It didn't allow for the fact that some option values
4614 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4615 the right test for positive values.
4617 48. Domain names in DNS records are case-independent. Exim always looks them up
4618 in lower case. Some resolvers return domain names in exactly the case they
4619 appear in the zone file, that is, they may contain uppercase letters. Not
4620 all resolvers do this - some return always lower case. Exim was treating a
4621 change of case by a resolver as a change of domain, similar to a widening
4622 of a domain abbreviation. This triggered its re-routing code and so it was
4623 trying to route what was effectively the same domain again. This normally
4624 caused routing to fail (because the router wouldn't handle the domain
4625 twice). Now Exim checks for this case specially, and just changes the
4626 casing of the domain that it ultimately uses when it transmits the message
4629 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4632 50. If a filter generated a file delivery with a non-absolute name (possible if
4633 no home directory exists for the router), the forbid_file option was not
4636 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4637 the existing equality matching.
4639 52. Exim was using ints instead of ino_t variables in some places where it was
4640 dealing with inode numbers.
4642 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4643 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4644 called TMPDIR, and if it finds it is different, it changes its value.
4646 54. The smtp_printf() function is now made available to local_scan() so
4647 additional output lines can be written before returning. There is also an
4648 smtp_fflush() function to enable the detection of a dropped connection.
4649 The variables smtp_input and smtp_batched_input are exported to
4652 55. Changed the default runtime configuration: the message "Unknown user"
4653 has been removed from the ACL, and instead placed on the localuser router,
4654 using the cannot_route_message feature. This means that any verification
4655 failures that generate their own messages won't get overridden. Similarly,
4656 the "Unrouteable address" message that was in the ACL for unverifiable
4657 relay addresses has also been removed.
4659 56. Added hosts_avoid_esmtp to the smtp transport.
4661 57. The exicyclog script was not checking for the esoteric option
4662 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4663 will work only if exicyclog is run under the appropriate euid.
4665 58. Following a discussion on the list, the rules by which Exim recognises line
4666 endings on incoming messages have been changed. The -dropcr and drop_cr
4667 options are now no-ops, retained only for backwards compatibility. The
4668 following line terminators are recognized: LF CRLF CR. However, special
4669 processing applies to CR:
4671 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4672 nor a local message in the state where . is a terminator.
4674 (ii) If a bare CR is encountered in a header line, an extra space is added
4675 after the line terminator so as not to end the header. The reasoning
4676 behind this is that bare CRs in header lines are most likely either
4677 to be mistakes, or people trying to play silly games.
4679 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4680 was being incorrectly given as 18 bytes larger than it should have been.
4681 This is a VOB (very old bug).
4683 60. This may never have affected anything current, but just in case it has:
4684 When the local host is found other than at the start of a list of hosts,
4685 the local host, those with the same MX, and any that follow, are discarded.
4686 When the list in question was part of a longer list of hosts, the following
4687 hosts (not currently being processed) were also being discarded. This no
4688 longer happens. I'm not sure if this situation could ever has previously
4691 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4694 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4695 header. This is something that is recommended in a new Internet Draft, and
4696 is something that is documented as being done by Sendmail. There are two
4697 possible values. For messages generated by the autoreply transport, Exim
4700 Auto-Submitted: auto-replied
4702 whereas for all other generated messages (e.g. bounces) it adds
4704 Auto-Submitted: auto-generated
4706 63. The "personal" condition in filters now includes a test for the
4707 Auto-Submitted: header. If it contains the string "auto-" the message it
4708 not considered personal.
4710 64. Added rcpt_include_affixes as a generic transport option.
4712 65. Added queue_only_override (default true).
4714 66. Added the syslog_duplication option.
4716 67. If what should have been the first header line of a message consisted of
4717 a space followed by a colon, Exim was mis-interpreting it as a header line.
4718 It isn't of course - it is syntactically invalid and should therefore be
4719 treated as the start of the message body. The misbehaviour could have
4720 caused a number of strange effects, including loss of data in subsequent
4721 header lines, and spool format errors.
4723 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4724 client host had authenticated. This control can now be exercised by an ACL
4725 for more flexibility.
4727 69. By default, callouts do not happen when testing with -bh. There is now a
4728 variant, -bhc, which does actually run the callout code, including
4729 consulting and updating the callout cache.
4731 70. Added support for saslauthd authentication, courtesy of Alexander
4734 71. If statvfs() failed on the spool or log directories while checking their
4735 size for availability, Exim confusingly gave the error "space shortage".
4736 Furthermore, in debugging mode it crashed with a floating point exception.
4737 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4738 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4739 serious problem, Exim now writes to the main and panic logs when this
4740 happens, with details of the failure. It then refuses to accept the
4741 incoming message, giving the message "spool directory problem" or "log
4742 directory problem" with a 421 code for SMTP messages.
4744 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4745 0, 1, and 2 exist, because some OS complain for execs without them (see
4746 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4747 descriptors. However, the code omitted to check that the open succeeded,
4748 causing mysterious errors if for some reason the permissions on /dev/null
4749 got screwed. Now Exim writes a message to the main and panic logs, and
4750 bombs out if it can't open /dev/null.
4752 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4753 interact so that it is all more flexible. It is supposed to remain
4754 backwards compatible. Also added extra_local_interfaces.
4756 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4757 to bomb out with an assertion failure - to the client this appears as a
4758 connection drop. This problem occurs in the part of the code that was taken
4759 from the Samba project. Fortunately, the assertion is in a very simple
4760 function, so I have fixed this by reproducing the function inline in the
4761 one place where it is called, and arranging for authentication to fail
4762 instead of killing the process with assert().
4764 75. The SPA client code was not working when the server requested OEM rather
4765 than Unicode encoding.
4767 76. Added code to make require_files with a specific uid setting more usable in
4768 the case where statting the file as root fails - usually a non-root-mounted
4769 NFS file system. When this happens and the failure is EACCES, Exim now
4770 forks a subprocess and does the per-uid checking as the relevant uid.
4772 77. Added process_log_path.
4774 78. If log_file_path was not explicitly set, a setting of check_log_space or
4775 check_log_inodes was ignored.
4777 79. If a space check for the spool or log partitions fails, the incident is now
4778 logged. Of course, in the latter case the data may get lost...
4780 80. Added the %p formatting code to string_format() so that it can be used to
4781 print addresses in debug_print(). Adjusted all the address printing in the
4782 debugging in store.c to use %p rather than %d.
4784 81. There was a concern that a line of code in smtp_in.c could overflow a
4785 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4786 initially expressed, the concern was not well-founded, because trailing
4787 spaces are removed early. However, if the trailing spaces were followed by
4788 a NULL, they did not get removed, so the overflow was possible. Two fixes
4791 (a) I re-wrote the offending code in a cleaner fashion.
4792 (b) If an incoming SMTP command contains a NULL character, it is rejected
4795 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4796 runs initgroups(), so that if the Exim user is in any additional groups,
4797 they will be used during message reception.
4803 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob