1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.524 2007/08/22 14:07:30 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
60 PH/07 The error message for a badly-placed control=no_multiline_responses left
61 "_responses" off the end of the name.
63 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
65 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
66 (without spaces) instead of just copying the configuration text.
68 PH/10 Added the /noupdate option to the ratelimit ACL condition.
70 PH/11 Added $max_received_linelength.
72 PH/12 Added +ignore_defer and +include_defer to host lists.
74 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
75 way in which PCRE > 7.0 is built.
77 PH/14 Implemented queue_only_load_latch.
79 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
80 MAIL command. The effect was to mangle the value on 64-bit systems.
82 PH/16 Another patch from the Sieve maintainer.
84 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
86 PH/18 If a system quota error occurred while trying to create the file for
87 a maildir delivery, the message "Mailbox is full" was not appended to the
88 bounce if the delivery eventually timed out. Change 4.67/27 below applied
89 only to a quota excession during the actual writing of the file.
95 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
96 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
99 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
100 issue a MAIL command.
102 PH/02 In an ACL statement such as
104 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
106 if a client was not listed at all, or was listed with a value other than
107 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
108 the condition was not true (as it should be), so access was not denied.
109 The bug was that the ! inversion was incorrectly passed on to the second
110 item. This has been fixed.
112 PH/03 Added additional dnslists conditions == and =& which are different from
113 = and & when the dns lookup returns more than one IP address.
115 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
116 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
118 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
119 FSYNC, which compiles an option called disable_fsync that allows for
120 bypassing fsync(). The documentation is heavily laced with warnings.
122 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
124 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
125 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
126 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
127 including adding "make clean"; (3) Added -fPIC when compiling the test
128 dynamically loaded module, to get rid of a warning.
130 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
131 message fails, move_frozen_messages = true and ignore_bounce_errors_after
132 = 0s. The bug is otherwise harmless.
134 PH/07 There was a bug in the dovecot authenticator such that the value of
135 $auth1 could be overwritten, and so not correctly preserved, after a
136 successful authentication. This usually meant that the value preserved by
137 the server_setid option was incorrect.
139 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
141 PH/09 Installed PCRE release 7.0.
143 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
144 run for batched SMTP input. It is now run at the start of every message
145 in the batch. While fixing this I discovered that the process information
146 (output by running exiwhat) was not always getting set for -bs and -bS
147 input. This is fixed, and it now also says "batched" for BSMTP.
149 PH/11 Added control=no_pipelining.
151 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
152 patch, slightly modified), and move the expansion of helo_data till after
153 the connection is made in the smtp transport (so it can use these
156 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
158 PH/14 Added log_selector = +pid.
160 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
162 PH/16 Add ${if forany and ${if forall.
164 PH/17 Added dsn_from option to vary the From: line in DSNs.
166 PH/18 Flush SMTP output before performing a callout, unless control =
167 no_callout_flush is set.
169 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
170 was true (the default) a successful delivery failed to delete the retry
171 item, thus causing premature timeout of the address. The bug is now
174 PH/20 Added hosts_avoid_pipelining to the smtp transport.
176 PH/21 Long custom messages for fakedefer and fakereject are now split up
177 into multiline reponses in the same way that messages for "deny" and
178 other ACL rejections are.
180 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
181 with slight modification.
183 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
184 draft, changing the syntax and factoring some duplicate code".
186 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
187 for deliveries of the second and subsequent messages over the same SMTP
190 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
191 ${reduce, with only minor "tidies".
193 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
195 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
196 expansion side effects.
198 PH/27 When a message times out after an over-quota error from an Exim-imposed
199 quota, the bounce message says "mailbox is full". This message was not
200 being given when it was a system quota that was exceeded. It now should
203 MH/03 Made $recipients available in local_scan(). local_scan() already has
204 better access to the recipient list through recipients_list[], but
205 $recipients can be useful in postmaster-provided expansion strings.
207 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
208 in the case of a MAIL command with additional options following the
209 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
210 were accidentally chopped off.
212 PH/29 SMTP synchronization checks are implemented when a command is read -
213 there is a check that no more input is waiting when there shouldn't be
214 any. However, for some commands, a delay in an ACL can mean that it is
215 some time before the response is written. In this time, more input might
216 arrive, invalidly. So now there are extra checks after an ACL has run for
217 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
218 pipelining has not been advertised.
220 PH/30 MH's patch to allow iscntrl() characters to be list separators.
222 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
223 returned in the SMTP response when smtp_return_error_details was false.
226 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
227 instead of the C I/O that was originally supplied, because problems were
230 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
231 Exim which did not show up earlier: it was assuming that a call to
232 SSL_CTX_set_info_callback() might give an error value. In fact, there is
233 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
234 was a macro that became an assignment, so it seemed to work. This has
235 changed to a proper function call with a void return, hence the compile
236 error. Exim's code has been fixed.
238 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
241 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
243 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
245 PH/37 If a message is not accepted after it has had an id assigned (e.g.
246 because it turns out to be too big or there is a timeout) there is no
247 "Completed" line in the log. When some messages of this type were
248 selected by exigrep, they were listed as "not completed". Others were
249 picked up by some special patterns. I have improved the selection
250 criteria to be more general.
252 PH/38 The host_find_failed option in the manualroute router can now be set
253 to "ignore", to completely ignore a host whose IP address cannot be
254 found. If all hosts are ignored, the behaviour is controlled by the new
255 host_all_ignored option.
257 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
258 homing or because of multiple MX records with /mx) generated more than
259 one IP address, and the following item turned out to be the local host,
260 all the secondary addresses of the first item were incorrectly removed
261 from the list, along with the local host and any following hosts (which
262 is what is supposed to happen).
264 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
265 whoever called Exim into the -H file. In the case of the daemon it was
266 behaving confusingly. When first started, it used values for whoever
267 started the daemon, but after a SIGHUP it used the Exim user (because it
268 calls itself on a restart). I have changed the code so that it now always
271 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
272 message are rejected with the same error (e.g. no authentication or bad
273 sender address), and a DATA command is nevertheless sent (as can happen
274 with PIPELINING or a stupid MUA), the error message that was given to the
275 RCPT commands is included in the rejection of the DATA command. This is
276 intended to be helpful for MUAs that show only the final error to their
279 PH/42 Another patch from the Sieve maintainer.
281 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
282 Eximstats - Fixed some broken HTML links and added missing column headers
284 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
285 columns for Rejects, Temp Rejects, Ham, and Spam rows.
287 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
289 PH/43 Yet another patch from the Sieve maintainer.
291 PH/44 I found a way to check for a TCP/IP connection going away before sending
292 the response to the final '.' that terminates a message, but only in the
293 case where the client has not sent further data following the '.'
294 (unfortunately, this is allowed). However, in many cases there won't be
295 any further data because there won't be any more messages to send. A call
296 to select() can be used: if it shows that the input is "ready", there is
297 either input waiting, or the socket has been closed. An attempt to read
298 the next input character can distinguish the two cases. Previously, Exim
299 would have sent an OK response which the client would never have see.
300 This could lead to message repetition. This fix should cure that, at
301 least in a lot of common cases.
303 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
304 advertised in response to EHLO.
310 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
311 fixed by 4.65/MH/01 (is this a record?) are fixed:
313 (i) An empty string was always treated as zero by the numeric comparison
314 operators. This behaviour has been restored.
316 (ii) It is documented that the numeric comparison operators always treat
317 their arguments as decimal numbers. This was broken in that numbers
318 starting with 0 were being interpreted as octal.
320 While fixing these problems I realized that there was another issue that
321 hadn't been noticed. Values of message_size_limit (both the global option
322 and the transport option) were treated as octal if they started with 0.
323 The documentation was vague. These values are now always treated as
324 decimal, and I will make that clear in the documentation.
330 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
331 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
334 MH/01 Don't check that the operands of numeric comparison operators are
335 integers when their expansion is in "skipping" mode (fixes bug
336 introduced by 4.64-PH/07).
338 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
339 child addresses, Exim now panics and dies. Previously, because the count
340 is held in a short int, deliveries were likely to be lost. As such a
341 large number of recipients for a single message is ridiculous
342 (performance will be very, very poor), I have chosen to impose a limit
343 rather than extend the field.
349 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
350 leftover -K file (the existence of which was triggered by #402).
351 While we were at it, introduced process PID as part of the -K
352 filename. This should rule out race conditions when creating
355 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
356 processing considerably. Previous code took too long for large mails,
357 triggering a timeout which in turn triggers #401.
359 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
360 in the DK code in transports.c. sendfile() is not really portable,
361 hence the _LINUX specificness.
363 TF/01 In the add_headers option to the mail command in an Exim filter,
364 there was a bug that Exim would claim a syntax error in any
365 header after the first one which had an odd number of characters
368 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
369 callout verification, Exim cached a "reject" for the entire domain. This
370 is correct for most verifications, but it is not correct for a recipient
371 verification with use_sender or use_postmaster set, because in that case
372 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
373 case of MAIL FROM:<> rejection from other early rejections (e.g.
374 rejection of HELO). When verifying a recipient using a non-null MAIL
375 address, the cache is ignored if it shows MAIL FROM:<> rejection.
376 Whatever the result of the callout, the value of the domain cache is
377 left unchanged (for any other kind of callout, getting as far as trying
378 RCPT means that the domain itself is ok).
380 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
383 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
384 manifest itself as EPIPE rather than ECONNECT. When tidying away a
385 session, the daemon ignores ECONNECT errors and logs others; it now
386 ignores EPIPE as well.
388 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
389 (quoted-printable decoding).
391 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
392 later the small subsequent patch to fix an introduced bug.
394 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
396 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
398 PH/08 An error is now given if message_size_limit is specified negative.
400 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
401 to be given (somewhat) arbitrary names.
403 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
406 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
407 miscellaneous code fixes
409 PH/10 Added the log_reject_target ACL modifier to specify where to log
412 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
413 hostname. This is wrong, because it relates to the incoming message (and
414 probably the interface on which it is arriving) and not to the outgoing
415 callout (which could be using a different interface). This has been
416 changed to use the value of the helo_data option from the smtp transport
417 instead - this is what is used when a message is actually being sent. If
418 there is no remote transport (possible with a router that sets up host
419 addresses), $smtp_active_hostname is used.
421 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
422 tweaks were necessary in order to get it to work (see also 21 below):
423 (a) The code assumed that strncpy() returns a negative number on buffer
424 overflow, which isn't the case. Replaced with Exim's string_format()
426 (b) There were several signed/unsigned issues. I just did the minimum
427 hacking in of casts. There is scope for a larger refactoring.
428 (c) The code used strcasecmp() which is not a standard C function.
429 Replaced with Exim's strcmpic() function.
430 (d) The code set only $1; it now sets $auth1 as well.
431 (e) A simple test gave the error "authentication client didn't specify
432 service in request". It would seem that Dovecot has changed its
433 interface. Fortunately there's a specification; I followed it and
434 changed what the client sends and it appears to be working now.
436 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
439 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
440 address A is aliased to B and C, where B exists and C does not. Without
441 -v the output is "A verified" because verification stops after a
442 successful redirection if more than one address is generated. However,
443 with -v the child addresses are also verified. Exim was outputting "A
444 failed to verify" and then showing the successful verification for C,
445 with its parentage. It now outputs "B failed to verify", showing B's
446 parentage before showing the successful verification of C.
448 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
449 look up a TXT record in a specific list after matching in a combined
452 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
453 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
454 they consult the DNS. I had assumed they would set it the way they
455 wanted; and indeed my experiments on Linux seem to show that in some
456 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
457 To be on the safe side, however, I have now made the interface to
458 host_find_byname() similar to host_find_bydns(), with an argument
459 containing the DNS resolver options. The host_find_byname() function now
460 sets these options at its start, just as host_find_bydns() does. The smtp
461 transport options dns_qualify_single and dns_search_parents are passed to
462 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
463 of host_find_byname() use the default settings of RES_DEFNAMES
464 (qualify_single) but not RES_DNSRCH (search_parents).
466 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
467 spool_read_header() do less string testing, by means of a preliminary
468 switch on the second character of optional "-foo" lines. (This is
469 overdue, caused by the large number of possibilities that now exist.
470 Originally there were few.) While I was there, I also converted the
471 str(n)cmp tests so they don't re-test the leading "-" and the first
472 character, in the hope this might squeeze out yet more improvement.
474 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
475 flag allowing group syntax was set by the header_syntax check but not
476 turned off, possible causing trouble later; (2) The flag was not being
477 set at all for the header_verify test, causing "group"-style headers to
478 be rejected. I have now set it in this case, and also caused header_
479 verify to ignore an empty address taken from a group. While doing this, I
480 came across some other cases where the code for allowing group syntax
481 while scanning a header line wasn't quite right (mostly, not resetting
482 the flag correctly in the right place). These bugs could have caused
483 trouble for malformed header lines. I hope it is now all correct.
485 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
486 with the "reply" argument non-NULL. The code, however (which originally
487 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
488 but it didn't always do it. This confused somebody who was copying the
489 code for some other use. I have removed all the tests.
491 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
492 feature that was used to support insecure browsers during the U.S. crypto
493 embargo. It requires special client support, and Exim is probably the
494 only MTA that supported it -- and would never use it because real RSA is
495 always available. This code has been removed, because it had the bad
496 effect of slowing Exim down by computing (never used) parameters for the
497 RSA_EXPORT functionality.
499 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
500 authenticator to fail if there's a tab character in the incoming data
501 (there should never be unless someone is messing about, as it's supposed
502 to be base64-encoded). Also added, on Timo's advice, the "secured" option
503 if the connection is using TLS or if the remote IP is the same as the
504 local IP, and the "valid-client-cert option" if a client certificate has
507 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
508 authenticators. This can be used for authorization after authentication
509 succeeds. (In the case of plaintext, it servers for both authentication
512 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
513 if any retry times were supplied.
515 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
516 connection, where there is no client IP address to check. In this
517 situation, the verify now always succeeds.
519 PH/25 Applied John Jetmore's -Mset patch.
521 PH/26 Added -bem to be like -Mset, but loading a message from a file.
523 PH/27 In a string expansion for a processed (not raw) header when multiple
524 headers of the same name were present, leading whitespace was being
525 removed from all of them, but trailing whitespace was being removed only
526 from the last one. Now trailing whitespace is removed from each header
527 before concatenation. Completely empty headers in a concatenation (as
530 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
531 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
533 PH/29 [Removed. This was a change that I later backed out, and forgot to
534 correct the ChangeLog entry (that I had efficiently created) before
535 committing the later change.]
537 PH/30 Exim was sometimes attempting to deliver messages that had suffered
538 address errors (4xx response to RCPT) over the same connection as other
539 messages routed to the same hosts. Such deliveries are always "forced",
540 so retry times are not inspected. This resulted in far too many retries
541 for the affected addresses. The effect occurred only when there were more
542 hosts than the hosts_max_try setting in the smtp transport when it had
543 the 4xx errors. Those hosts that it had tried were not added to the list
544 of hosts for which the message was waiting, so if all were tried, there
545 was no problem. Two fixes have been applied:
547 (i) If there are any address or message errors in an SMTP delivery, none
548 of the hosts (tried or untried) are now added to the list of hosts
549 for which the message is waiting, so the message should not be a
550 candidate for sending over the same connection that was used for a
551 successful delivery of some other message. This seems entirely
552 reasonable: after all the message is NOT "waiting for some host".
553 This is so "obvious" that I'm not sure why it wasn't done
554 previously. Hope I haven't missed anything, but it can't do any
555 harm, as the worst effect is to miss an optimization.
557 (ii) If, despite (i), such a delivery is accidentally attempted, the
558 routing retry time is respected, so at least it doesn't keep
559 hammering the server.
561 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
562 in ${readsocket because some servers need this prod.
564 PH/32 Added some extra debug output when updating a wait-xxx database.
566 PH/33 The hint "could be header name not terminated by colon", which has been
567 given for certain expansion errors for a long time, was not being given
568 for the ${if def:h_colon_omitted{... case.
570 PH/34 The spec says: "With one important exception, whenever a domain list is
571 being scanned, $domain contains the subject domain." There was at least
572 one case where this was not true.
574 PH/35 The error "getsockname() failed: connection reset by peer" was being
575 written to the panic log as well as the main log, but it isn't really
576 panic-worthy as it just means the connection died rather early on. I have
577 removed the panic log writing for the ECONNRESET error when getsockname()
580 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
581 runs only) independently of the message's sender address. This meant
582 that, if the 4xx error was in fact related to the sender, a different
583 message to the same recipient with a different sender could confuse
584 things. In particualar, this can happen when sending to a greylisting
585 server, but other circumstances could also provoke similar problems.
586 I have changed the default so that the retry time for these errors is now
587 based a combination of the sender and recipient addresses. This change
588 can be overridden by setting address_retry_include_sender=false in the
591 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
592 remote server are returned as part of bounce messages. This was not
593 happening for LMTP over a pipe (the lmtp transport), but now it is the
594 same for both kinds of LMTP.
596 PH/38 Despite being documented as not happening, Exim was rewriting addresses
597 in header lines that were in fact CNAMEs. This is no longer the case.
599 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
600 and queue runs started by the daemon processed all messages. This has
601 been fixed so that -R and -S can now usefully be given with -q<time>.
603 PH/40 Import PCRE release 6.7 (fixes some bugs).
605 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
607 PH/42 Give an error if -q is specified more than once.
609 PH/43 Renamed the variables $interface_address and $interface_port as
610 $received_ip_address and $received_port, to make it clear that these
611 values apply to message reception, and not to the outgoing interface when
612 a message is delivered. (The old names remain recognized, of course.)
614 PH/44 There was no timeout on the connect() call when using a Unix domain
615 socket in the ${readsocket expansion. There now is.
617 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
618 be meaningful with "accept".
620 SC/01 Eximstats V1.43
621 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
623 SC/02 Eximstats V1.44
624 Use a glob alias rather than an array ref in the generated
625 parser. This improves both readability and performance.
627 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
628 Collect SpamAssassin and rejection statistics.
629 Don't display local sender or destination tables unless
630 there is data to show.
631 Added average volumes into the top table text output.
633 SC/04 Eximstats V1.46
634 Collect data on the number of addresses (recipients)
635 as well as the number of messages.
637 SC/05 Eximstats V1.47
638 Added 'Message too big' to the list of mail rejection
639 reasons (thanks to Marco Gaiarin).
641 SC/06 Eximstats V1.48
642 Mainlog lines which have GMT offsets and are too short to
643 have a flag are now skipped.
645 SC/07 Eximstats V1.49 (Alain Williams)
646 Added the -emptyok flag.
648 SC/08 Eximstats V1.50
649 Fixes for obtaining the IP address from reject messages.
651 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
652 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
653 whitesspace changes from 4.64-PH/27
655 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
658 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
659 are found, allow negative numbers in numeric criteria)
661 JJ/06 exipick.20061117.2, added new $message_body_missing variable
663 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
664 to match changes made in 4.64-PH/43
666 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
668 PH/47 Put in an explicit test for a DNS lookup of an address record where the
669 "domain" is actually an IP address, and force a failure. This locks out
670 those revolvers/nameservers that support "A-for-A" lookups, in
671 contravention of the specifications.
673 PH/48 When a host name was looked up from an IP address, and the subsequent
674 forward lookup of the name timed out, the host name was left in
675 $sender_host_name, contrary to the specification.
677 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
678 restricted to single-key lookups, Exim was not diagnosing an error if
679 * or *@ was used with a query-style lookup.
681 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
683 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
684 long ago, but noone interested enough thought of it. Let's just say that
685 the "1.1" means that there are some new functions that weren't there at
686 some point in the past.
688 PH/51 Error processing for expansion failure of helo_data from an smtp
689 transport during callout processing was broken.
691 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
692 tested/used via the -bh/-bhc/-bs options.
694 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
695 bug, fixed in subsequent PCRE releases).
697 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
698 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
700 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
706 SC/01 Use a glob alias rather than an array ref in eximstats generated
707 parser. This improves both readability and performance.
709 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
710 Don't display local sender or destination tables in eximstats unless
711 there is data to show.
712 Added average volumes into the eximstats top table text output.
714 SC/03 Collect data on the number of addresses (recipients) as well
715 as the number of messages in eximstats.
717 TF/01 Correct an error in the documentation for the redirect router. Exim
718 does (usually) call initgroups() when daemonizing.
720 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
721 with consistent privilege compared to when running as a daemon.
723 TF/03 Note in the spec that $authenticated_id is not set for local
724 submissions from trusted users.
726 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
727 Thanks to Dean Brooks <dean@iglou.com> for the patch.
729 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
730 by adding some example configuration directives to the default
731 configuration file. A little bit of work is required to uncomment the
732 directives and define how usernames and passwords are checked, but
733 there is now a framework to start from.
735 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
736 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
737 without this. I don't know how relevant this is to other LDAP libraries.
739 PH/02 Add the verb name to the "unknown ACL verb" error.
741 PH/03 Magnus Holmgren's patch for filter_prepend_home.
743 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
745 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
746 directory not expanded when it should be if an expanded home directory
747 was set for the address (which is overridden by the transport).
749 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
752 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
753 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
754 because it is too late at that time, and has no effect.
756 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
757 security issue with \' (bugzilla #107). I could not use the
758 PQescapeStringConn() function, because it needs a PGconn value as one of
761 PH/08 When testing addresses using -bt, indicate those final addresses that
762 are duplicates that would not cause an additional delivery. At least one
763 person was confused, thinking that -bt output corresponded to deliveries.
764 (Suppressing duplicates isn't a good idea as you lose the information
765 about possibly different redirections that led to the duplicates.)
767 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
768 systems where poll() doesn't work, in particular OS X.
770 PH/10 Added more information to debugging output for retry time not reached.
772 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
773 operations in malware.c.
775 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
778 PH/13 If write_rejectlog was set false when logging was sent to syslog with
779 syslog_duplication set false, log lines that would normally be written
780 both the the main log and to the reject log were not written to syslog at
783 PH/14 In the default configuration, change the use of "message" in ACL warn
784 statements to "add_header".
786 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
787 not followed by a command (e.g. "seen endif").
789 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
790 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
793 PH/17 Added extra conditions to the default value of delay_warning_condition
797 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
798 { match{$h_precedence:}{(?i)bulk|list|junk} } \
799 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
802 The Auto-Submitted: and various List- headers are standardised, whereas I
803 don't think Precedence: ever was.
805 PH/18 Refactored debugging code in route_finduser() to show more information,
806 in particular, the error code if getpwnam() issues one.
808 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
809 This is apparently needed in addition to the PH/07 change above to avoid
810 any possible encoding problems.
812 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
813 but not after initializing Perl.
815 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
816 output them only if debugging. By default they are written stderr,
817 apparently, which is not desirable.
819 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
822 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
825 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
827 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
828 authenticated or an ident call has been made. Suppress the default
829 values for $authenticated_id and $authenticated_sender (but permit -oMai
830 and -oMas) when testing with -bh.
832 PH/24 Re-jigged the order of the tests in the default configuration so that the
833 tests for valid domains and recipients precede the DNS black list and CSA
834 tests, on the grounds that those ones are more expensive.
836 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
837 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
838 This bug exists in every version of Exim that I still have, right back to
841 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
842 However, an attempt to turn on case-sensitivity in a regex key by
843 including (?-i) didn't work because the subject string was already
844 lowercased, and the effects were non-intuitive. It turns out that a
845 one-line patch can be used to allow (?-i) to work as expected.
851 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
852 other effects) broke the use of negated acl sub-conditions.
854 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
857 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
858 "Deny" causes Exim to reject the incoming connection with a 554 error.
859 Unfortunately, if there is a major crisis, such as a disk failure,
860 tcp-wrappers gives "deny", whereas what one would like would be some
861 kind of temporary error. A kludge has been added to help with this.
862 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
863 554 error is used if errno is still zero or contains ENOENT (which occurs
864 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
867 PH/03 Add -lutil to the default FreeBSD LIBS setting.
869 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
870 errors. Otherwise a message that provokes a temporary error (when other
871 messages do not) can cause a whole host to time out.
873 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
874 addresses were routed directly to files or pipes from a redirect router.
875 File deliveries just didn't batch; pipe deliveries might have suffered
878 PH/06 A failure to get a lock for a hints database would erroneously always say
879 "Failed to get write lock", even when it was really a read lock.
881 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
882 of 0600. This has been changed to use the value of the lockfile_mode
883 option (which defaults to 0600).
885 PH/08 Applied small patch from the Sieve maintainer.
887 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
888 folder from quota calculations, a direct delivery into this folder messed
889 up the contents of the maildirsize file. This was because the regex was
890 used only to exclude .Trash (or whatever) when the size of the mailbox
891 was calculated. There was no check that a delivery was happening into an
892 excluded directory. This bug has been fixed by ignoring all quota
893 processing for deliveries into excluded directories.
895 PH/10 Added the maildirfolder_create_regex option to appendfile.
901 PH/01 The code for finding all the local interface addresses on a FreeBSD
902 system running IPv6 was broken. This may well have applied to all BSD
903 systems, as well as to others that have similar system calls. The broken
904 code found IPv4 interfaces correctly, but gave incorrect values for the
905 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
906 that it would not match correctly against @[] and not recognize the IPv6
909 PH/02 The ipliteral router was not recognizing addresses of the form user@
910 [ipv6:....] because it didn't know about the "ipv6:" prefix.
912 PH/03 Added disable_ipv6.
914 PH/04 Changed $reply_address to use the raw form of the headers instead of the
915 decoded form, because it is most often used to construct To: headers
916 lines in autoreplies, and the decoded form may well be syntactically
917 invalid. However, $reply_address has leading white space removed, and all
918 newlines turned into spaces so that the autoreply transport does not
921 PH/05 If group was specified without a user on a router, and no group or user
922 was specified on a transport, the group from the router was ignored.
924 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
925 for visible compile-time settings that can be used to change these
926 numbers, for those that want even more. Backwards compatibility with old
927 spool files has been maintained. However, going back to a previous Exim
928 release will lost any variables that are in spool files.
930 PH/07 Two small changes when running in the test harness: increase delay when
931 passing a TCP/IP connection to a new process, in case the original
932 process has to generate a bounce, and remove special handling of
933 127.0.0.2 (sic), which is no longer necessary.
935 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
936 be the same on different OS.
938 PH/09 Moved a debug statement in filter processing to avoid a race problem when
941 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
942 whether --show-vars was specified or not
944 JJ/02 exipick: Added support for new ACL variable spool format introduced
947 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
948 syntactically invalid From: or Reply-to: line, and a filter used this to
949 generate an autoreply, and therefore failed to obtain an address for the
950 autoreply, Exim could try to deliver to a non-existent relative file
951 name, causing unrelated and misleading errors. What now happens is that
952 it logs this as a hard delivery error, but does not attempt to create a
955 PH/11 The exinext utility has a -C option for testing purposes, but although
956 the given file was scanned by exinext itself; it wasn't being passed on
957 when Exim was called.
959 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
960 an end-of-file indication when reading a command response.
962 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
963 compiled. In many other places in Exim, IPv6 addresses are always
964 recognized, so I have changed this. It also means that IPv4 domain
965 literals of the form [IPV4:n.n.n.n] are now always recognized.
967 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
968 used if the router is not running as root, for example, when verifying at
969 ACL time, or when using -bh. The debugging output from this situation was
970 non-existent - all you got was a failure to exec. I have made two
973 (a) Failures to set uid/gid, the current directory, or a process leader
974 in a subprocess such as that created by queryprogram now generate
975 suitable debugging ouput when -d is set.
977 (b) The queryprogram router detects when it is not running as root,
978 outputs suitable debugging information if -d is set, and then runs
979 the subprocess without attempting to change uid/gid.
981 PH/15 Minor change to Makefile for building test_host (undocumented testing
984 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
985 additional section of a DNS packet that returns MX or SRV records.
986 Instead, it always explicitly searches for A/AAAA records. This avoids
987 major problems that occur when a DNS server includes only records of one
988 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
989 fixed another bug: if SRV records were looked up and the corresponding
990 address records were *not* found in the additional section, the port
991 values from the SRV records were lost.
993 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
994 using the correct key (the original address) when searching the retry
995 rules in order to find which one to use for generating the retry hint.
997 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
998 adding the default one. Similarly, if it contains a Reply-To: header, the
999 errors_reply_to option, if set, is not used.
1001 PH/19 When calculating a retry time, Exim used to measure the "time since
1002 failure" by looking at the "first failed" field in the retry record. Now
1003 it does not use this if it is later than than the arrival time of the
1004 message. Instead it uses the arrival time. This makes for better
1005 behaviour in cases where some deliveries succeed, thus re-setting the
1006 "first failed" field. An example is a quota failure for a huge message
1007 when small messages continue to be delivered. Without this change, the
1008 "time since failure" will always be short, possible causing more frequent
1009 delivery attempts for the huge message than are intended.
1010 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1012 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1013 $1, $2, $3) because the numerical variables can be reset during some
1014 expansion items (e.g. "match"), thereby losing the authentication data.
1016 PH/21 Make -bV show the size of off_t variables so that the test suite can
1017 decide whether to run tests for quotas > 2G.
1019 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1020 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1021 filecount value is greater than 2G or if a quota value is greater than 2G
1022 on a system where the size of off_t is not greater than 4, a panic error
1025 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1026 never match. The debug and -bh output now contains an explicit error
1027 message indicating a malformed IPv4 address or mask.
1029 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1030 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1031 PH/23 above applies.
1033 PH/25 Do not write to syslog when running in the test harness. The only
1034 occasion when this arises is a failure to open the main or panic logs
1035 (for which there is an explicit test).
1037 PH/26 Added the /no_tell option to "control=freeze".
1039 PH/27 If a host name lookup failed very early in a connection, for example, if
1040 the IP address matched host_lookup and the reverse lookup yielded a name
1041 that did not have a forward lookup, an error message of the form "no IP
1042 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1043 could be logged. Now it outputs the IP address instead of "NULL".
1045 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1046 allows the sender and the authenticated sender to be set when
1047 submitting a message from within Exim. Since child_open_exim() is
1048 documented for local_scan(), the new function should be too.
1050 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1051 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1052 results in an empty string is now treated as unset.
1054 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1056 PH/31 Added sender_verify_fail logging option.
1058 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1059 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1060 tidied the source and removed it altogether.
1062 PH/33 When a queue run was abandoned because the load average was too high, a
1063 log line was always written; now it is written only if the queue_run log
1064 selector is set. In addition, the log line for abandonment now contains
1065 information about the queue run such as the pid. This is always present
1066 in "start" and "stop" lines but was omitted from the "abandon" line.
1068 PH/34 Omit spaces between a header name and the colon in the error message that
1069 is given when verify = headers_syntax fails (if there are lots of them,
1070 the message gets confusing).
1072 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1073 names, as there are now some PTR records that contain slashes. This check
1074 is only to protect against broken name servers that fall over on strange
1075 characters, so the fact that it applies to all lookups doesn't matter.
1077 PH/36 Now that the new test suite is complete, we can remove some of the
1078 special code in Exim that was needed for the old test suite. For example,
1079 sorting DNS records because real resolvers return them in an arbitrary
1080 order. The new test suite's fake resolver always returns records in the
1083 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1084 bounces) except when queue_only is set, to avoid logging races between
1085 the different processes.
1087 PH/38 Panic-die if .include specifies a non-absolute path.
1089 PH/39 A tweak to the "H" retry rule from its user.
1091 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1092 a label. They prevented compilation on older perls.
1094 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1095 a warning to be raised on newish perls.
1097 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1098 on queue. Changes to match documented behaviour of showing count of
1099 messages matching specified criteria.
1101 PH/40 Changed the default ident timeout from 30s to 5s.
1103 PH/41 Added support for the use of login_cap features, on those BSD systems
1104 that have them, for controlling the resources used by pipe deliveries.
1106 PH/42 The content-scanning code uses fopen() to create files in which to put
1107 message data. Previously it was not paying any attention to the mode of
1108 the files. Exim runs with umask(0) because the rest of the code creates
1109 files with open(), and sets the required mode explicitly. Thus, these
1110 files were ending up world-writeable. This was not a big issue, because,
1111 being within the spool directory, they were not world-accessible. I have
1112 created a function called modefopen, which takes an additional mode
1113 argument. It sets umask(777), creates the file, chmods it to the required
1114 mode, then resets the umask. All the relevant calls to fopen() in the
1115 content scanning code have been changed to use this function.
1117 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1118 to 24 hours. This avoids potential overflow problems when processing G
1119 and H retry rules. I suspect nobody ever tinkers with this value.
1121 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1123 PH/45 When the plaintext authenticator is running as a client, the server's
1124 challenges are checked to ensure they are valid base64 strings. By
1125 default, the authentication attempt is cancelled if an invalid string is
1126 received. Setting client_ignore_invalid_base64 true ignores these errors.
1127 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1128 they are received. Thus, the responses can be made to depend on the
1129 challenges. If an invalid string is ignored, an empty string is placed in
1132 PH/46 Messages that are created by the autoreply transport now contains a
1133 References: header, in accordance with RFCs 2822 and 3834.
1135 PH/47 Added authenticated_sender_force to the smtp transport.
1137 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1139 PH/49 Installed latest patch from the Sieve maintainer.
1141 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1142 was also set, the appendfile transport was unnecessarily scanning a
1143 directory of message files (e.g. for maildir delivery) to find the count
1144 of files (along with the size), even though it did not need this
1145 information. It now does the scan only if it needs to find either the
1146 size of the count of files.
1148 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1150 PH/52 Two bugs concerned with error handling when the smtp transport is
1153 (i) Exim was not creating retry information for temporary errors given
1154 for individual recipients after the DATA command when the smtp transport
1155 was used in LMTP mode. This meant that they could be retried too
1156 frequently, and not timed out correctly.
1158 (ii) Exim was setting the flag that allows error details to be returned
1159 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1160 recipients that were returned after the DATA command.
1162 PH/53 This is related to PH/52, but is more general: for any failing address,
1163 when detailed error information was permitted to be returned to the
1164 sender, but the error was temporary, then after the final timeout, only
1165 "retry timeout exceeded" was returned. Now it returns the full error as
1166 well as "retry timeout exceeded".
1168 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1169 do this, and (what is worse) MTAs that accept it.
1171 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1172 will now be deprecated.
1174 PH/56 New os.c-cygwin from the Cygwin maintainer.
1176 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1177 formats (previously only available in exim formats via -bpr, -bpru,
1178 and -bpra. Now also available in native and exiqgrep formats)
1180 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1181 with very large, slow to parse queues
1183 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1185 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1187 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1188 responses to authentication challenges, though it was showing the
1189 challenges; (ii) I've removed the CR characters from the debug output for
1192 PH/58 Allow for the insertion of a newline as well as a space when a string
1193 is turned into more than one encoded-word during RFC 2047 encoding. The
1194 Sieve code now uses this.
1196 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1197 data_4xx, lost_connection, tls_required.
1199 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1200 message was being sent as an SMTP response.
1202 PH/61 Add -l and -k options to exicyclog.
1204 PH/62 When verifying, if an address was redirected to one new address, so that
1205 verification continued, and the new address failed or deferred after
1206 having set something in $address_data, the value of $address_data was not
1207 passed back to the ACL. This was different to the case when no
1208 redirection occurred. The value is now passed back in both cases.
1210 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1211 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1212 use login_cap.h, so on its own it isn't the distinguishing feature. The
1213 new name refers directly to the setclassresources() function.
1215 PH/65 Added configuration files for NetBSD3.
1217 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1219 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1220 is preferred over IPv4.
1222 PH/68 The bounce_return_message and bounce_return_body options were not being
1223 honoured for bounces generated during the reception of non-SMTP messages.
1224 In particular, this applied to messages rejected by the ACL. This bug has
1225 been fixed. However, if bounce_return_message is true and bounce_return_
1226 body is false, the headers that are returned for a non-SMTP message
1227 include only those that have been read before the error was detected.
1228 (In the case of an ACL rejection, they have all been read.)
1230 PH/69 The HTML version of the specification is now built in a directory called
1231 spec_html instead of spec.html, because the latter looks like a path with
1232 a MIME-type, and this confuses some software.
1234 PH/70 Catch two compiler warnings in sieve.c.
1236 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1237 function verify_get_ident() calls ip_connect() to connect a socket, but
1238 if the "connect()" function timed out, ip_connect() used to close the
1239 socket. However, verify_get_ident() also closes the socket later, and in
1240 between Exim writes to the log, which may get opened at this point. When
1241 the socket was closed in ip_connect(), the log could get the same file
1242 descriptor number as the socket. This naturally causes chaos. The fix is
1243 not to close the socket in ip_connect(); the socket should be closed by
1244 the function that creates it. There was only one place in the code where
1245 this was missing, in the iplookup router, which I don't think anybody now
1246 uses, but I've fixed it anyway.
1248 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1249 well as to direct DNS lookups. Otherwise the handling of names in host
1250 lists is inconsistent and therefore confusing.
1256 PH/01 Two changes to the default runtime configuration:
1258 (1) Move the checks for relay_from_hosts and authenticated clients from
1259 after to before the (commented out) DNS black list checks.
1261 (2) Add control=submission to the relay_from_hosts and authenticated
1262 clients checks, on the grounds that messages accepted by these
1263 statements are most likely to be submissions.
1265 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1267 (1) Generate an error if the third argument for the ${prvs expansion is
1270 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1273 (3) Reset the variables that are obtained from the first argument of
1274 ${prvscheck and used in the second argument before leaving the code,
1275 because their memory is reclaimed, so using them afterwards may do
1278 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1279 one (it's much easier than Tom thought :-).
1281 (5) Because of (4), we can now allow for the use of $prvscheck_result
1282 inside the third argument.
1284 PH/03 For some reason, the default setting of PATH when running a command from
1285 a pipe transport was just "/usr/bin". I have changed it to
1288 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1289 anything to be listed in the output from -bV.
1291 PH/05 When a filter generated an autoreply, the entire To: header line was
1292 quoted in the delivery log line, like this:
1294 => >A.N.Other <ano@some.domain> <original@ddress> ...
1296 This has been changed so that it extracts the operative address. There
1297 may be more than one such address. If so, they are comma-separated, like
1300 => >ano@some.domain,ona@other.domain <original@ddress> ...
1302 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1303 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1304 not being looked up in the rDNS to get a host name, Exim was showing the
1305 IP address twice in Received: lines, even though the IP addresses were
1306 identical. For example:
1308 Received: from [1.2.3.4] (helo=[1.2.3.4])
1310 However, if the real host name was known, it was omitting the HELO data
1311 if it matched the actual IP address. This has been tidied up so that it
1312 doesn't show the same IP address twice.
1314 PH/07 When both +timestamp and +memory debugging was on, the value given by
1315 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1316 called by the debug printing, thereby overwriting the timestamp buffer.
1317 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1319 PH/08 When the original message was included in an autoreply transport, it
1320 always said "this is a copy of the message, including all the headers",
1321 even if body_only or headers_only was set. It now gives an appropriate
1324 PH/09 Applied a patch from the Sieve maintainer which:
1326 o fixes some comments
1327 o adds the (disabled) notify extension core
1328 o adds some debug output for the result of if/elsif tests
1329 o points to the current vacation draft in the documentation
1330 and documents the missing references header update
1334 o fixes a bug in processing the envelope test (when testing
1335 multiple envelope elements, the last element determinted the
1338 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1339 Electronic Mail") by including:
1341 Auto-submitted: auto-generated
1343 in the messages that it generates (bounce messages and others, such as
1344 warnings). In the case of bounce messages for non-SMTP mesages, there was
1345 also a typo: it was using "Auto_submitted" (underscore instead of
1346 hyphen). Since every message generated by Exim is necessarily in response
1347 to another message, thes have all been changed to:
1349 Auto-Submitted: auto-replied
1351 in accordance with these statements in the RFC:
1353 The auto-replied keyword:
1355 - SHOULD be used on messages sent in direct response to another
1356 message by an automatic process,
1358 - MUST NOT be used on manually-generated messages,
1360 - MAY be used on Delivery Status Notifications (DSNs) and Message
1361 Disposition Notifications (MDNs),
1363 - MUST NOT be used on messages generated by automatic or periodic
1364 processes, except for messages which are automatic responses to
1367 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1368 to the default Received: header definition.
1370 PH/12 Added log selector acl_warn_skipped (default on).
1372 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1373 variables because (a) they are in the wrong storage pool and (b) even if
1374 they were copied, it wouldn't work properly because of the caching.
1376 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1377 checking when decoding. Apparently there are clients that generate
1378 overlong encoded strings. Why am I not surprised?
1380 PH/15 If the first argument of "${if match_address" was not empty, but did not
1381 contain an "@" character, Exim crashed. Now it writes a panic log message
1382 and treats the condition as false.
1384 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1386 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1387 extension "envelope-auth". The code is finished and in agreement with
1388 other implementations, but there is no documentation so far and in fact,
1389 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1390 not changing the active code.
1392 Print executed "if" and "elsif" statements when debugging is used. This
1393 helps a great deal to understand what a filter does.
1395 Document more things not specified clearly in RFC3028. I had all this
1396 sorted out, when out of a sudden new issues came to my mind. Oops."
1398 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1401 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1402 canonical form (as documented). However, after a host name lookup from
1403 the IP address, check_host() was doing a simple string comparison with
1404 addresses acquired from the DNS when checking that the found name did
1405 have the original IP as one of its addresses. Since any found IPv6
1406 addresses are likely to be in abbreviated form, the comparison could
1407 fail. Luckily, there already exists a function for doing the comparison
1408 by converting both addresses to binary, so now that is used instead of
1409 the text comparison.
1411 PH/20 There was another similar case to PH/19, when a complete host name was
1412 given in a host list; looking up its IP address could give an abbreviated
1413 form, whereas the current host's name might or might not be abbreviated.
1414 The same fix has been applied.
1420 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1421 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1424 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1425 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1427 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1429 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1430 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1431 Also, for "Auto-Submitted", treat anything other than "no" as
1432 non-personal, in accordance with RFC 3834. (Previously it treated
1433 anything starting "auto-" as non-personal.)
1435 TF/01 The control=submission/name=... option had a problem with syntax
1436 errors if the name included a slash character. The /name= option
1437 now slurps the rest of the string, so it can include any characters
1438 but it must come last in the list of options (after /sender_retain
1441 PH/05 Some modifications to the interface to the fake nameserver for the new
1449 TK/01 Added the "success_on_redirect" address verification option. See
1450 NewStuff for rationale and an example.
1452 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1454 PH/02 Patch to exigrep to allow it to work on syslog lines.
1456 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1457 fread() to skip over the body file's header line, because in Cygwin the
1458 header line is locked and is inaccessible.
1460 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1461 co-exist for some time) to make it clear that it is the Exim ID that is
1462 referenced, not the Message-ID: header line.
1464 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1465 string_format() function, because snprintf() does not exist on all
1468 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1469 ${stat: expansion item.
1471 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1472 protocol synchronization error", to keep the pedants happy.
1474 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1475 well as for IRIX systems, when gcc is being used. See the host.c source
1478 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1480 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1483 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1484 transport and to the smtp transport in LMTP mode.
1486 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1488 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1489 run a filter in a subprocess. This could lead to confusion in subsequent
1490 lookups in the parent process. There should also be a search_tidyup() at
1491 the end of the subprocess.
1493 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1494 only if the host matched helo_try_verify_hosts, which caused the
1495 verification to occur when the EHLO/HELO command was issued. The ACL just
1496 tested the remembered result. Now, if a previous verification attempt has
1497 not happened, "verify = helo" does it there and then.
1499 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1501 TK/03 Fix log output including CR from clamd.
1503 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1504 exist provoked a memory error which could cause a segfault.
1506 PH/15 Installed PCRE 6.2
1508 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1510 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1511 of the problem. Specifically, suggested +O2 rather than +O1 for the
1514 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1516 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1517 record, the port was not correctly logged when the outgoing_port log
1518 selector was set (it logged the transort's default port).
1520 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1521 fallback_hosts, and "hosts" in the smtp transport.
1523 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1524 host errors such as "Connection refused".
1526 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1527 authentication with radiusclient 0.4.9:
1529 - Error returned from rc_read_config was caught wrongly
1530 - Username/password not passed on to radius server due to wrong length.
1532 The presumption is that some radiusclient API changes for 4.51/PH/17
1533 were not taken care of correctly. The code is still untested by me (my
1534 Linux distribution still has 0.3.2 of radiusclient), but it was
1535 contributed by a Radius user.
1537 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1538 expanding the "port" option of the smtp transport.
1540 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1541 while reading a MIME header. Thanks to Tom Hughes for a patch.
1543 PH/24 Include config.h inside local_scan.h so that configuration settings are
1546 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1547 that in an ACL for RCPT (for example), you can examine exactly what was
1550 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1551 commands, but it was not correctly comparing the address with the actual
1552 client host address. Thus, it would show the EHLO address in Received:
1553 header lines when this was not necessary.
1555 PH/27 Added the % operator to ${eval:}.
1557 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1558 it should be ignoring failures (because with -C, for example, it has lost
1559 privilege). It wasn't ignoring creation failures other than "already
1562 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1565 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1566 because an input line was too long, either on its own, or by virtue of
1567 too many continuations, the temporary file was not being removed, and the
1568 return code was incorrect.
1570 PH/31 Missing "BOOL" in function definition in filtertest.c.
1572 PH/32 Applied Sieve patches from the maintainer.
1574 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1576 PH/33 Added "verify = not_blind".
1578 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1579 Local/Makefile (with some defaults set). These are used in built scripts
1580 such as exicyclog, but they have never been used in the exim_install
1581 script (though there are many overriding facilities there). I have
1582 arranged that the exim_install script now takes note of these two
1585 PH/35 Installed configuration files for Dragonfly.
1587 PH/36 When a locally submitted message by a trusted user did not contain a
1588 From: header, and the sender address was obtained from -f or from an SMTP
1589 MAIL command, and the trusted user did not use -F to supply a sender
1590 name, $originator_name was incorrectly used when constructing a From:
1591 header. Furthermore, $originator_name was used for submission mode
1592 messages from external hosts without From: headers in a similar way,
1593 which is clearly wrong.
1595 PH/37 Added control=suppress_local_fixups.
1597 PH/38 When log_selector = +received_sender was set, and the addition of the
1598 sender made the log line's construction buffer exactly full, or one byte
1599 less than full, an overflow happened when the terminating "\n" was
1602 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1603 when the result of a list match is failure because a DNS lookup failed.
1605 PH/40 RM_COMMAND is now used in the building process.
1607 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1608 the "build-* directories that it finds.
1610 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1611 address in a domain literal was a prefix of an interface address.
1613 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1614 when verifying a sender address, unless rewrite_headers is false.
1616 PH/44 Wrote a long comment about why errors_to addresses are verified as
1617 recipients, not senders.
1619 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1620 the ratelimit ACL was added.
1622 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1624 PH/46 Added extra information about PostgreSQL errors to the error string.
1626 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1627 avoiding the need to install special zones in a real server. This is
1628 backwards compatible; if it can't find the fake resolver, it drops back.
1629 Thus, both old and new test suites can be run.
1631 TF/02 Added util/ratelimit.pl
1633 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1634 clock is set back in time.
1636 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1637 Candler <B.Candler@pobox.com>.
1639 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1640 OK for addresses that are the result of redirections.
1642 PH/48 A number of further additions for the benefit of the new test suite,
1643 including a fake gethostbyname() that interfaces to the fake DNS resolver
1646 TF/06 The fix for widen_domains has also been applied to qualify_single and
1647 search_parents which are the other dnslookup options that can cause
1650 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1653 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1655 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1656 service" if it encountered an MX record with an empty target hostname.
1657 The message is now "an MX or SRV record indicated no SMTP service".
1659 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1660 if the DNS of the sending site is misconfigured. This is quite a
1661 common situation. This change restores the behaviour of treating a
1662 helo verification defer as a failure.
1664 PH/51 If self=fail was set on a router, the bounce message did not include the
1665 actual error message.
1671 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1673 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1674 command itself ended in error, the underlying message about the transport
1675 filter timeout was being overwritten with the pipe command error. Now the
1676 underlying error message should be appended to the second error message.
1678 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1680 PH/02 Reduce the amount of output that "make" produces by default. Full output
1681 can still be requested.
1683 PH/03 The warning log line about a condition test deferring for a "warn" verb
1684 was being output only once per connection, rather than after each
1685 occurrence (because it was using the same function as for successful
1686 "warn" verbs). This seems wrong, so I have changed it.
1688 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1689 it should not have, which might have caused a crash in the right
1690 circumstances, but probably never did.
1692 PH/04 Installed a modified version of Tony Finch's patch to make submission
1693 mode fix the return path as well as the Sender: header line, and to
1694 add a /name= option so that you can make the user's friendly name appear
1697 TF/03 Added the control = fakedefer ACL modifier.
1699 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1700 Mark Lowes for thorough testing.
1702 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1704 TK/03 Merged latest SRS patch from Miles Wilton.
1706 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1707 with the definition in sysexits.h (which is #included earlier).
1708 Fortunately, Exim does not actually use EX_OK. The code used to try to
1709 preserve the sysexits.h value, by assumimg that macro definitions were
1710 scanned for macro replacements. I have been disabused of this notion,
1711 so now the code just undefines EX_OK before #including unistd.h.
1713 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1714 in the smtp transport. When a block could not be written in a single
1715 write() function, the timeout was being re-applied to each part-write.
1716 This seems wrong - if the receiver was accepting one byte at a time it
1717 would take for ever. The timeout is now adjusted when this happens. It
1718 doesn't have to be particularly precise.
1720 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1721 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1723 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1724 without a domain if the check to <postmaster@domain> fails.
1726 SC/01 Eximstats: added -xls and the ability to specify output files
1727 (patch written by Frank Heydlauf).
1729 SC/02 Eximstats: use FileHandles for outputing results.
1731 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1733 SC/04 Eximstats: fixed display of large numbers with -nvr option
1735 SC/05 Eximstats: fixed merging of reports with empty tables.
1737 SC/06 Eximstats: added the -include_original_destination flag
1739 SC/07 Eximstats: removed tabs and trailing whitespace.
1741 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1743 TK/06 MBOX spool code: Add real "From " MBOX separator line
1744 so the .eml file is really in mbox format (even though
1745 most programs do not really care). Patch from Alex Miller.
1747 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1748 The latter is generated from $received_to and is only set if the
1749 message has one envelope recipient. SA can use these headers,
1750 obviously out-of-the-box. Patch from Alex Miller.
1752 PH/08 The ${def test on a variable was returning false if the variable's
1753 value was "0", contrary to what the specification has always said!
1754 The result should be true unless the variable is empty.
1756 PH/09 The syntax error of a character other than { following "${if
1757 def:variable_name" (after optional whitespace) was not being diagnosed.
1758 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1759 accidental colon was present, for example, could give incorrect results.
1761 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1762 result is used (not including appendfile, where other changes are about
1765 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1766 This involved changing a lot of size variables from int to off_t. It
1767 should work with maildirs and everything.
1769 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1770 spamd dying while we are connected to it.
1772 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1775 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1776 function for BDB 4.3.
1778 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1780 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1781 the sources for PCRE have been re-arranged and more files are now
1784 PH/15 The code I had for printing potentially long long variables in PH/11
1785 above was not the best (it lost precision). The length of off_t variables
1786 is now inspected at build time, and an appropriate printing format (%ld
1787 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1788 to be "long long int" or "long int". This is needed for the internal
1789 formatting function string_vformat().
1791 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1792 the configuration file to be ":syslog", then the script "guesses" where
1793 the logs files are, rather than using the compiled in default. In our
1794 case the guess is not the same as the compiled default, so the script
1795 suddenly stopped working when I started to use syslog. The patch checks
1796 to see if log_file_path is "". If so, it attempts to read it from exim
1797 with no configuration file to get the compiled in version, before it
1798 falls back to the previous guessing code."
1800 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1801 implementing BATV in an Exim configuration. See NewStuff for the gory
1804 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1805 Makefile that are specific to HP-UX.
1807 PH/18 If the "use_postmaster" option was set for a recipient callout together
1808 with the "random" option, the postmaster address was used as the MAIL
1809 FROM address for the random test, but not for the subsequent recipient
1810 test. It is now used for both.
1812 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1813 patch removes a few documentation additions to RFC 3028, because the
1814 latest draft now contains them. It adds the new en;ascii-case comparator
1815 and a new error check for 8bit text in MIME parts. Comparator and
1816 require names are now matched exactly. I enabled the subaddress
1817 extension, but it is not well tested yet (read: it works for me)."
1819 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1820 rework some of the code of TK/09 above to avoid the hardwired use of
1821 "%lld" and "long long". Replaced the call to snprintf() with a call to
1824 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1825 records point to non-existent hosts", "retry timeout exceeded", and
1826 "retry time not reached for any host after a long failure period".
1828 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1829 experimental DomainKeys support:
1831 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1832 (2) On an error such as an illegally used "control", the wrong name for
1833 the control was given.
1835 These problems did NOT occur unless DomainKeys support was compiled.
1837 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1839 PH/24 Added ${if match_ip condition.
1841 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1842 databases so that it will be absolutely obvious if a crash occurs in the
1843 DB library. This is a regular occurrence (often caused by mis-matched
1846 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1847 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1848 picked up on a user's system that detects such things. There doesn't seem
1849 to be a gcc warning option for this - only an attribute that has to be
1850 put on the function's prototype. It seems that in Fedora Core 4 they have
1851 set this on a number of new functions. No doubt there will be more in due
1854 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1855 specify a transport. However, if an address that was verified by such a
1856 router was the subject of a callout, Exim crashed because it tried to
1857 read the rcpt_include_affixes from the non-existent transport. Now it
1858 just assumes that the setting of that option is false. This bug was
1859 introduced by 4.51/PH/31.
1861 PH/28 Changed -d+all to exclude +memory, because that information is very
1862 rarely of interest, but it makes the output a lot bigger. People tend to
1863 do -d+all out of habit.
1865 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1866 code in os-type was giving problems when libc.so lives in lib64, like on
1869 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1870 aren't the modern standard, and it seems that some systems' include files
1871 don't always have them. Exim was already checking for some of the newer
1872 ones like T_AAAA, and defining it itself. I've added checks for all the
1873 record types that Exim uses.
1875 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1876 not automatically generating a new one, as it is supposed to. This
1877 prevented TLS from working. If the file did exist, but contained invalid
1878 data, a new version was generated, as expected. It was only the case of a
1879 non-existent file that was broken.
1881 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1882 with a change in libdomainkeys > 0.64.
1884 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1885 from DNS. If the selector record carries the flag, it now has
1886 precedence over the domain-wide flag.
1888 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1890 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1891 the use of a port name that isn't defined in /etc/services) occurred, the
1892 message was deferred as in a normal delivery, and thus remained on the
1893 spool, instead of being failed because of the mua_wrapper setting. This
1894 is now fixed, and I tidied up some of the mua_wrapper messages at the
1897 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1898 the messages in a hash of arrays rather than using individual hashes.
1899 This is a bit cleaner and results in dramatic memory savings, albeit
1900 at a slight CPU cost.
1902 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1903 as requested by Marc Sherman.
1905 SC/10 Eximstats: added histograms for user specified patterns as requested
1908 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1910 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1911 fopen() in the content-scanning modules that did not already have it.
1917 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1918 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1920 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1922 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1924 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1925 to test Sieve filters that use "vacation".
1927 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1928 that changes the way the GnuTLS parameters are stored in the cache file.
1929 The new format can be generated externally. For backward compatibility,
1930 if the data in the cache doesn't make sense, Exim assumes it has read an
1931 old-format file, and it generates new data and writes a new file. This
1932 means that you can't go back to an older release without removing the
1935 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1936 work if there are any delivery delays because "one_time" forces the
1937 parent to be marked "delivered", so its unseen clone is never tried
1938 again. For this reason, Exim now forbids the simultaneous setting of
1941 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1942 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1943 entry if you want to know the details. The fix, however, neglected to
1944 consider the case when local delivery batching is involved. The test for
1945 "previously delivered" was not happening when checking to see if an
1946 address could be batched with a previous (undelivered) one; under
1947 certain circumstances this could lead to multiple deliveries to the same
1950 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1951 in its include files, and this causes problems building Exim.
1953 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1954 header_syntax) but Exim was just ignoring anything given after a slash.
1955 In particular, this caused confusion with an attempt to use "verify =
1956 reverse_host_lookup/defer_ok". An error is now given when options are
1957 supplied for verify items that do not have them. (Maybe reverse_host_
1958 lookup should have a defer_ok option, but that's a different point.)
1960 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1961 defined by RFC 821) to 2048, because there were problems with some AUTH
1962 commands, and RFC 1869 says the size should be increased for extended
1963 SMTP commands that take arguments.
1965 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1968 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1969 "unknown" error; now it says that the functionality isn't in the binary.
1971 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1972 an address' error message when a string expansion fails (syntax or
1973 whatever). Otherwise the password may appear in the log. Following change
1974 PH/42 below, there is no longer a chance of it appearing in a bounce
1977 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1979 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1980 of its addresses was ever tried. (Bugzilla bug #2.)
1982 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1983 the result incorrectly in the debug output. (It correctly added a newline
1984 to what was transported.)
1986 TF/01 Added $received_time.
1988 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1989 commented out examples of how to interface to a virus scanner and to
1990 SpamAssassin. Also added commented examples of av_scanner and
1991 spamd_address settings.
1993 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1994 and controls are allowed in which ACLs. There were a couple of minor
1995 errors. Some of the entries in the conditions table (which is a table of
1996 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1997 negation of where the condition IS allowed.
1999 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2001 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2002 header file does not have a version number, so I've had to invent a new
2003 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2004 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2005 radiusclient), but it was contributed by a Radius user.
2007 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2008 files or directories, for OpenSSL.
2010 PH/19 When an Exim process that is running as root has to create an Exim log
2011 file, it does so in a subprocess that runs as exim:exim so as to get the
2012 ownership right at creation (otherwise, other Exim processes might see
2013 the file with the wrong ownership). There was no test for failure of this
2014 fork() call, which would lead to the process getting stuck as it waited
2015 for a non-existent subprocess. Forks do occasionally fail when resources
2016 run out. I reviewed all the other calls to fork(); they all seem to check
2019 PH/20 When checking for unexpected SMTP input at connect time (before writing
2020 the banner), Exim was not dealing correctly with a non-positive return
2021 from the read() function. If the client had disconnected by this time,
2022 the result was a log entry for a synchronization error with an empty
2023 string after "input=" when read() returned zero. If read() returned -1
2024 (an event I could not check), uninitialized data bytes were printed.
2025 There were reports of junk text (parts of files, etc) appearing after
2028 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2030 PH/22 Added support for macro redefinition, and (re)definition in between
2031 driver and ACL definitions.
2033 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2034 forgetting to use the resulting value; it was using the unexpanded value.
2036 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2037 hadn't been configured. The fix is from Juergen Kreileder, who
2038 understands it better than I do:
2040 "Here's what I see happening with three configured cyrus_sasl
2041 authenticators configured (plain, login, cram-md5):
2043 On startup auth_cyrus_sasl_init() gets called for each of these.
2044 This means three calls to sasl_listmech() without a specified mech_list.
2045 => SASL tests which mechs of all available mechs actually work
2046 => three warnings about OTP not working
2047 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2049 With the patch, sasl_listmech() also gets called three times. But now
2050 SASL's mech_list option is set to the server_mech specified in the the
2051 authenticator. Or in other words, the answer from sasl_listmech()
2052 gets limited to just the mech you're testing for (which is different
2054 => the return list contains just 'plain' or 'login', 'cram-md5' or
2055 nothing depending on the value of ob->server_mech.
2057 I've just tested the patch: Authentication still works fine,
2058 unavailable mechs specified in the exim configuration are still
2059 caught, and the auth.log warnings about OTP are gone."
2061 PH/25 When debugging is enabled, the contents of the command line are added
2062 to the debugging output, even when log_selector=+arguments is not
2065 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2066 answer is "GNU", and only if the return is "GNU/something" is the answer
2069 PH/27 $acl_verify_message is now set immediately after the failure of a
2070 verification in an ACL, and so is available in subsequent modifiers. In
2071 particular, the message can be preserved by coding like this:
2073 warn !verify = sender
2074 set acl_m0 = $acl_verify_message
2076 Previously, $acl_verify_message was set only while expanding "message"
2077 and "log_message" when a very denied access.
2079 PH/28 Modified OS/os.c-Linux with
2081 -#ifndef OS_LOAD_AVERAGE
2082 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2084 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2085 nomenclature these days.)
2087 PH/29 Installed patch from the Sieve maintainer that adds the options
2088 sieve_useraddress and sieve_subaddress to the redirect router.
2090 PH/30 In these circumstances:
2091 . Two addresses routed to the same list of hosts;
2092 . First host does not offer TLS;
2093 . First host accepts first address;
2094 . First host gives temporary error to second address;
2095 . Second host offers TLS and a TLS session is established;
2096 . Second host accepts second address.
2097 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2098 and peerdn, if requested) that were in fact used only for the second
2101 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2102 attention to any local part prefix or suffix that was matched by the
2103 router that accepted the address. It now behaves in the same way as it
2104 does for delivery: the affixes are removed from the local part unless
2105 rcpt_include_affixes is set on the transport.
2107 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2108 timeout during the DATA phase of an incoming message.
2110 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2111 applied a patch sanctioned by the Sieve maintainer.
2113 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2114 the uid or gid is negative. A case of a negative gid caused this to be
2115 noticed. The fix allows for either to be negative.
2117 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2118 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2121 PH/36 The change PH/12 above was broken. Fixed it.
2123 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2124 the grounds that routing the same address twice would always produce the
2125 same answer. This might have been true once, but it is certainly no
2126 longer true now. Routing a child address may depend on the previous
2127 routing that produced that child. Some complicated redirection strategies
2128 went wrong when messages had multiple recipients, and made Exim's
2129 behaviour dependent on the order in which the addresses were given.
2131 I have moved the duplicate checking until after the routing is complete.
2132 Exim scans the addresses that are assigned to local and remote
2133 transports, and removes any duplicates. This means that more work will be
2134 done, as duplicates will always all be routed, but duplicates are
2135 presumably rare, so I don't expect this is of any significance.
2137 For deliveries to pipes, files, and autoreplies, the duplicate checking
2138 still happens during the routing process, since they are not going to be
2141 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2142 It corrects a timeout issue with spamd. This is Ian's comment: "The
2143 background is that sometimes spamd either never reads data from a
2144 connection it has accepted, or it never writes response data. The exiscan
2145 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2146 blindly assumes that writes won't block so it may never time out."
2148 PH/39 Allow G after quota size as well as K and M.
2150 PH/40 The value set for $authenticated_id in an authenticator may not contain
2151 binary zeroes or newlines because the value is written to log lines and
2152 to spool files. There was no check on this. Now the value is run through
2153 the string_printing() function so that such characters are converted to
2154 printable escape sequences.
2156 PH/41 $message_linecount is a new variable that contains the total number of
2157 lines in the message. Compare $body_linecount, which is the count for the
2160 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2161 bounce and delay warning messages, except in certain special cases, which
2164 (a) An SMTP error message from a remote host;
2165 (b) A message specified in a :fail: redirection;
2166 (c) A message specified in a "fail" command in a system filter;
2167 (d) A message specified in a FAIL return from the queryprogram router;
2168 (e) A message specified by the cannot_route_message router option.
2170 In these cases only, Exim does include the error details in bounce and
2171 warning messages. There are also a few cases where bland messages such
2172 as "unrouteable address" or "local delivery error" are given.
2174 PH/43 $value is now also set for the "else" part of a ${run expansion.
2176 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2177 being worked on, but at least Exim now implements the latest version to
2180 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2181 process to complete was treated as a delivery failure, a timeout while
2182 writing the message to the pipe was logged, but erroneously treated as a
2183 successful delivery. Such timeouts include transport filter timeouts. For
2184 consistency with the overall process timeout, these timeouts are now
2185 treated as errors, giving rise to delivery failures by default. However,
2186 there is now a new Boolean option for the pipe transport called
2187 timeout_defer, which, if set TRUE, converts the failures into defers for
2188 both kinds of timeout. A transport filter timeout is now identified in
2191 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2192 systems where "make" and "gmake" are different, calling "gmake" at top
2193 level broke things. I've arranged for the value of $(MAKE) to be passed
2194 from the Makefile to this script so that it can call the same version of
2198 A note about Exim versions 4.44 and 4.50
2199 ----------------------------------------
2201 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2202 changes of various kinds. As a consequence, a big documentation update was
2203 needed. This delayed the release for rather longer than seemed good, especially
2204 in the light of a couple of (minor) security issues. Therefore, the changes
2205 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2206 release. So 4.44 and 4.50 are in effect two different branches that both start
2209 I have left the 4.50 change log unchanged; it contains all the changes since
2210 4.43. The change log for 4.44 is below; many of its items are identical to
2211 those for 4.50. This seems to be the most sensible way to preserve the
2212 historical information.
2218 1. Minor wording change to the doc/README.SIEVE file.
2220 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2221 computation of the current number of files was incorrect.
2223 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2224 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2227 4. Give more explanation in the error message when the command for a transport
2228 filter fails to execute.
2230 5. There are several places where Exim runs a non-Exim command in a
2231 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2232 was being done only for the command run by the queryprogram router. It is
2233 now done for all such subprocesses. The other cases are: ${run, transport
2234 filters, and the commands run by the lmtp and pipe transports.
2236 6. Added CONFIGURE_GROUP build-time option.
2238 7. Some older OS have a limit of 256 on the maximum number of file
2239 descriptors. Exim was using setrlimit() to set 1000 as a large value
2240 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2241 systems. I've change it so that if it can't get 1000, it tries for 256.
2243 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2244 was an oversight, and furthermore, ever since the addition of extra
2245 controls (e.g. 4.43/32), the checks on when to allow different forms of
2246 "control" were broken. There should now be diagnostics for all cases when a
2247 control that does not make sense is encountered.
2249 9. Added the /retain_sender option to "control=submission".
2251 10. $recipients is now available in the predata ACL (oversight).
2253 11. Tidy the search cache before the fork to do a delivery from a message
2254 received from the command line. Otherwise the child will trigger a lookup
2255 failure and thereby defer the delivery if it tries to use (for example) a
2256 cached ldap connection that the parent has called unbind on.
2258 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2259 of $address_data from the recipient verification was clobbered by the
2260 sender verification.
2262 13. The value of address_data from a sender verification is now available in
2263 $sender_address_data in subsequent conditions in the ACL statement.
2265 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2267 15. Added a new option "connect=<time>" to callout options, to set a different
2270 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2271 was its contents. (It was OK if the option was not defined at all.)
2273 17. A "Completed" log line is now written for messages that are removed from
2274 the spool by the -Mrm option.
2276 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2277 information about exactly what failed.
2279 19. Added -dd to debug only the daemon process.
2281 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2282 handles timeouts, both on the server side and network timeouts. Renamed the
2283 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2285 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2286 It is now set to "smtps".
2288 22. $host_address is now set to the target address during the checking of
2289 ignore_target_hosts.
2291 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2292 being passed; this would have caused $sender_host_name to have been used if
2293 matching the list had actually called for a host name (not very likely,
2294 since this list is usually IP addresses). A host name is now passed as
2297 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2298 code that specifies a non-restarting handler (typically sigaction() in
2299 modern systems) in an attempt to fix a rare and obscure crash bug.
2301 25. Narrowed the window for a race in the daemon that could cause it to ignore
2302 SIGCHLD signals. This is not a major problem, because they are used only to
2303 wake it up if nothing else does.
2305 26. A malformed maildirsize file could cause Exim to calculate negative values
2306 for the mailbox size or file count. Odd effects could occur as a result.
2307 The maildirsize information is now recalculated if the size or filecount
2310 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2311 support for a long time. Removed HAVE_SYS_VFS_H.
2313 28. Installed the latest version of exipick from John Jetmore.
2315 29. In an address list, if the pattern was not a regular expression, an empty
2316 subject address (from a bounce message) matched only if the pattern was an
2317 empty string. Non-empty patterns were not even tested. This was the wrong
2318 because it is perfectly reasonable to use an empty address as part of a
2319 database query. An empty address is now tested by patterns that are
2320 lookups. However, all the other forms of pattern expect the subject to
2321 contain a local part and a domain, and therefore, for them, an empty
2322 address still always fails if the pattern is not itself empty.
2324 30. Exim went into a mad DNS loop when attempting to do a callout where the
2325 host was specified on an smtp transport, and looking it up yielded more
2326 than one IP address.
2328 31. Re-factored the code for checking spool and log partition space into a
2329 function that finds that data and another that does the check. The former
2330 is then used to implement four new variables: $spool_space, $log_space,
2331 $spool_inodes, and $log_inodes.
2333 32. The RFC2047 encoding function was originally intended for short strings
2334 such as real names; it was not keeping to the 75-character limit for
2335 encoded words that the RFC imposes. It now respects the limit, and
2336 generates multiple encoded words if necessary. To be on the safe side, I
2337 have increased the buffer size for the ${rfc2047: expansion operator from
2340 33. It is now permitted to omit both strings after an "if" condition; if the
2341 condition is true, the result is "true". As before, when the second string
2342 is omitted, a false condition yields an empty string. This makes it less
2343 cumbersome to write custom ACL and router conditions.
2345 34. Failure to deliver a bounce message always caused it to be frozen, even if
2346 there was an errors_to setting on the router. The errors_to setting is now
2349 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2350 canonical form (fully expanded) before being placed in
2351 $sender_host_address.
2353 36. The table in the code that translates DNS record types into text (T_A to
2354 "A" for instance) was missing entries for NS and CNAME. It is just possible
2355 that this could have caused confusion if both these types were looked up
2356 for the same domain, because the text type is used as part of Exim's
2357 per-process caching. But the chance of anyone hitting this buglet seems
2360 37. The dnsdb lookup has been extended in a number of ways.
2362 (1) There is a new type, "zns", which walks up the domain tree until it
2363 finds some nameserver records. It should be used with care.
2365 (2) There is a new type, "mxh", which is like "mx" except that it returns
2366 just the host names, not the priorities.
2368 (3) It is now possible to give a list of domains (or IP addresses) to be
2369 looked up. The behaviour when one of the lookups defers can be
2370 controlled by a keyword.
2372 (4) It is now possible to specify the separator character for use when
2373 multiple records are returned.
2375 38. The dnslists ACL condition has been extended: it is now possible to supply
2376 a list of IP addresses and/or domains to be looked up in a particular DNS
2379 39. Added log_selector=+queue_time_overall.
2381 40. When running the queue in the test harness, wait just a tad after forking a
2382 delivery process, to get repeatability of debugging output.
2384 41. Include certificate and key file names in error message when GnuTLS fails
2385 to set them up, because the GnuTLS error message doesn't include the name
2386 of the failing file when there is a problem reading it.
2388 42. Allow both -bf and -bF in the same test run.
2390 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2392 44. The "Exiscan patch" is now merged into the mainline Exim source.
2394 45. Sometimes the final signoff response after QUIT could fail to get
2395 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2396 < 0 before doing a fflush(). This bug looks as though it goes back to the
2397 introduction of TLS in release 3.20, but "sometimes" must have been rare
2398 because the tests only now provoked it.
2400 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2401 (this can affect the format of dates).
2403 47. exim_tidydb, when checking for the continued existence of a message for
2404 which it has found a message-specific retry record, was not finding
2405 messages that were in split spool directories. Consequently, it was
2406 deleting retry records that should have stayed in existence.
2408 48. Steve fixed some bugs in eximstats.
2410 49. The SPA authentication driver was not abandoning authentication and moving
2411 on to the next authenticator when an expansion was forced to fail,
2412 contradicting the general specification for all authenticators. Instead it
2413 was generating a temporary error. It now behaves as specified.
2415 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2416 (the order specifies the preference for clients). The order is now AES256,
2417 AES128, 3DES, ARCFOUR128.
2419 51. Small patch to Sieve code - explicitly set From: when generating an
2422 52. Exim crashed if a remote delivery caused a very long error message to be
2423 recorded - for instance if somebody sent an entire SpamAssassin report back
2424 as a large number of 550 error lines. This bug was coincidentally fixed by
2425 increasing the size of one of Exim's internal buffers (big_buffer) that
2426 happened as part of the Exiscan merge. However, to be on the safe side, I
2427 have made the code more robust (and fixed the comments that describe what
2430 53. Now that there can be additional text after "Completed" in log lines (if
2431 the queue_time_overall log selector is set), a one-byte patch to exigrep
2432 was needed to allow it to recognize "Completed" as not the last thing in
2435 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2436 patch that reportedly fixes this has been added. I am not expert enough to
2437 create a test for it. This is what the patch creator wrote:
2439 "I found a little strange behaviour of ldap code when working with
2440 Windows 2003 AD Domain, where users was placed in more than one
2441 Organization Units. When I tried to give exim partial DN, the exit code
2442 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2443 But simultaneously result of request was absolutely normal ldap result,
2444 so I produce this patch..."
2446 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2447 REFERENCE, so I have modified the code to exclude the patch when that macro
2450 55. Some experimental protocols are using DNS PTR records for new purposes. The
2451 keys for these records are domain names, not reversed IP addresses. The
2452 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2453 leaves it alone. Component reversal etc. now happens only for IP addresses.
2456 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2458 57. Double the size of the debug message buffer (to 2048) so that more of very
2459 long debug lines gets shown.
2461 58. The exicyclog utility now does better if the number of log files to keep
2462 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2464 59. Two changes related to the smtp_active_hostname option:
2466 (1) $smtp_active_hostname is now available as a variable.
2467 (2) The default for smtp_banner uses $smtp_active_hostname instead
2468 of $primary_hostname.
2470 60. The host_aton() function is supposed to be passed a string that is known
2471 to be a valid IP address. However, in the case of IPv6 addresses, it was
2472 not checking this. This is a hostage to fortune. Exim now panics and dies
2473 if the condition is not met. A case was found where this could be provoked
2474 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2475 components; fortuitously, this particular loophole had already been fixed
2476 by change 4.50/55 above.
2478 If there are any other similar loopholes, the new check in host_aton()
2479 itself should stop them being exploited. The report I received stated that
2480 data on the command line could provoke the exploit when Exim was running as
2481 exim, but did not say which command line option was involved. All I could
2482 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2483 running as the user.
2486 61. There was a buffer overflow vulnerability in the SPA authentication code
2487 (which came originally from the Samba project). I have added a test to the
2488 spa_base64_to_bits() function which I hope fixes it.
2491 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2492 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2494 63. The daemon start-up calls getloadavg() while still root for those OS that
2495 need the first call to be done as root, but it missed one case: when
2496 deliver_queue_load_max is set with deliver_drop_privilege. This is
2497 necessary for the benefit of the queue runner, because there is no re-exec
2498 when deliver_drop_privilege is set.
2500 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2501 This has been fixed.
2503 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2504 was in use, was not putting the data itself into the right store pool;
2505 consequently, it could be overwritten for a subsequent message in the same
2506 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2509 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2511 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2512 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2513 the function were treating the return as a boolean value, which happened to
2514 work because 0=false and not-0=true, but is not correct code.
2516 68. The host_aton() function was not handling scoped IPv6 addresses (those
2517 with, for example, "%eth0" on the end) correctly.
2519 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2520 negated items (that is, ~something) in unsigned ints. Some compilers
2521 apparently mutter when there is no cast.
2523 70. If an address verification called from an ACL failed, and did not produce a
2524 user-specific message (i.e. there was only a "system" message), nothing was
2525 put in $acl_verify_message. In this situation, it now puts the system
2528 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2529 session; change 4.31/43 added the unwanted input to the log line - except
2530 that it did not do this in the start of session case. It now does.
2532 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2533 This is wrong and can cause the other end to generate a synchronization
2534 error if it is another Exim or anything else that does the synchronization
2535 check. A QUIT command is no longer sent after a timeout.
2537 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2538 during host lookups.
2540 74. The defer_ok option of callout verification was not working if it was used
2541 when verifying addresses in header lines, that is, for this case:
2543 verify = header_sender/callout=defer_ok
2545 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2546 those file descriptors could be used for SMTP connections. If anything
2547 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2548 could be sent to the SMTP client, causing chaos. The daemon now opens
2549 stdin, stdout, and stderr to /dev/null when it puts itself into the
2552 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2553 log by default. The user can override this with suitable Perl magic.
2555 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2556 the log message when discard triggers, was not working for the DATA ACL or
2557 for the non-SMTP ACL.
2559 78. Error message wording change in sieve.c.
2561 79. If smtp_accept_max_per_host was set, the number of connections could be
2562 restricted to fewer than expected, because the daemon was trying to set up
2563 a new connection before checking whether the processes handling previous
2564 connections had finished. The check for completed processes is now done
2565 earlier. On busy systems, this bug wouldn't be noticed because something
2566 else would have woken the daemon, and it would have reaped the completed
2569 80. If a message was submitted locally by a user whose login name contained one
2570 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2571 It caused a spool format error. I have fixed the spool reading code. A
2572 related problem was that the "from" clause in the Received: line became
2573 illegal because of the space(s). It is now covered by ${quote_local_part.
2575 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2578 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2580 83. Patch from Sieve maintainer to fix unterminated string problem in
2581 "vacation" handling.
2583 84. Some minor changes to the Linux configuration files to help with other
2584 OS variants using glibc.
2586 85. One more patch for Sieve to update vacation handling to latest spec.
2589 ----------------------------------------------------
2590 See the note above about the 4.44 and 4.50 releases.
2591 ----------------------------------------------------
2597 1. Change 4.43/35 introduced a bug that caused file counts to be
2598 incorrectly computed when quota_filecount was set in an appendfile
2601 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2602 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2605 3. Give more explanation in the error message when the command for a transport
2606 filter fails to execute.
2608 4. There are several places where Exim runs a non-Exim command in a
2609 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2610 was being done only for the command run by the queryprogram router. It is
2611 now done for all such subprocesses. The other cases are: ${run, transport
2612 filters, and the commands run by the lmtp and pipe transports.
2614 5. Some older OS have a limit of 256 on the maximum number of file
2615 descriptors. Exim was using setrlimit() to set 1000 as a large value
2616 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2617 systems. I've change it so that if it can't get 1000, it tries for 256.
2619 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2620 was an oversight, and furthermore, ever since the addition of extra
2621 controls (e.g. 4.43/32), the checks on when to allow different forms of
2622 "control" were broken. There should now be diagnostics for all cases when a
2623 control that does not make sense is encountered.
2625 7. $recipients is now available in the predata ACL (oversight).
2627 8. Tidy the search cache before the fork to do a delivery from a message
2628 received from the command line. Otherwise the child will trigger a lookup
2629 failure and thereby defer the delivery if it tries to use (for example) a
2630 cached ldap connection that the parent has called unbind on.
2632 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2633 of $address_data from the recipient verification was clobbered by the
2634 sender verification.
2636 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2637 was its contents. (It was OK if the option was not defined at all.)
2639 11. A "Completed" log line is now written for messages that are removed from
2640 the spool by the -Mrm option.
2642 12. $host_address is now set to the target address during the checking of
2643 ignore_target_hosts.
2645 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2646 being passed; this would have caused $sender_host_name to have been used if
2647 matching the list had actually called for a host name (not very likely,
2648 since this list is usually IP addresses). A host name is now passed as
2651 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2652 code that specifies a non-restarting handler (typically sigaction() in
2653 modern systems) in an attempt to fix a rare and obscure crash bug.
2655 15. Narrowed the window for a race in the daemon that could cause it to ignore
2656 SIGCHLD signals. This is not a major problem, because they are used only to
2657 wake it up if nothing else does.
2659 16. A malformed maildirsize file could cause Exim to calculate negative values
2660 for the mailbox size or file count. Odd effects could occur as a result.
2661 The maildirsize information is now recalculated if the size or filecount
2664 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2665 support for a long time. Removed HAVE_SYS_VFS_H.
2667 18. Updated exipick to current release from John Jetmore.
2669 19. Allow an empty sender to be matched against a lookup in an address list.
2670 Previously the only cases considered were a regular expression, or an
2673 20. Exim went into a mad DNS lookup loop when doing a callout where the
2674 host was specified on the transport, if the DNS lookup yielded more than
2677 21. The RFC2047 encoding function was originally intended for short strings
2678 such as real names; it was not keeping to the 75-character limit for
2679 encoded words that the RFC imposes. It now respects the limit, and
2680 generates multiple encoded words if necessary. To be on the safe side, I
2681 have increased the buffer size for the ${rfc2047: expansion operator from
2684 22. Failure to deliver a bounce message always caused it to be frozen, even if
2685 there was an errors_to setting on the router. The errors_to setting is now
2688 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2689 canonical form (fully expanded) before being placed in
2690 $sender_host_address.
2692 24. Updated eximstats to version 1.33
2694 25. Include certificate and key file names in error message when GnuTLS fails
2695 to set them up, because the GnuTLS error message doesn't include the name
2696 of the failing file when there is a problem reading it.
2698 26. Expand error message when OpenSSL has problems setting up cert/key files.
2701 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2702 (this can affect the format of dates).
2704 28. exim_tidydb, when checking for the continued existence of a message for
2705 which it has found a message-specific retry record, was not finding
2706 messages that were in split spool directories. Consequently, it was
2707 deleting retry records that should have stayed in existence.
2709 29. eximstats updated to version 1.35
2710 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2711 1.35 - bugfix such that pie charts by volume are generated correctly
2713 30. The SPA authentication driver was not abandoning authentication and moving
2714 on to the next authenticator when an expansion was forced to fail,
2715 contradicting the general specification for all authenticators. Instead it
2716 was generating a temporary error. It now behaves as specified.
2718 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2719 (the order specifies the preference for clients). The order is now AES256,
2720 AES128, 3DES, ARCFOUR128.
2722 31. Small patch to Sieve code - explicitly set From: when generating an
2725 32. Exim crashed if a remote delivery caused a very long error message to be
2726 recorded - for instance if somebody sent an entire SpamAssassin report back
2727 as a large number of 550 error lines. This bug was coincidentally fixed by
2728 increasing the size of one of Exim's internal buffers (big_buffer) that
2729 happened as part of the Exiscan merge. However, to be on the safe side, I
2730 have made the code more robust (and fixed the comments that describe what
2733 33. Some experimental protocols are using DNS PTR records for new purposes. The
2734 keys for these records are domain names, not reversed IP addresses. The
2735 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2736 leaves it alone. Component reversal etc. now happens only for IP addresses.
2739 34. The host_aton() function is supposed to be passed a string that is known
2740 to be a valid IP address. However, in the case of IPv6 addresses, it was
2741 not checking this. This is a hostage to fortune. Exim now panics and dies
2742 if the condition is not met. A case was found where this could be provoked
2743 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2744 components; fortuitously, this particular loophole had already been fixed
2745 by change 4.50/55 or 4.44/33 above.
2747 If there are any other similar loopholes, the new check in host_aton()
2748 itself should stop them being exploited. The report I received stated that
2749 data on the command line could provoke the exploit when Exim was running as
2750 exim, but did not say which command line option was involved. All I could
2751 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2752 running as the user.
2755 35. There was a buffer overflow vulnerability in the SPA authentication code
2756 (which came originally from the Samba project). I have added a test to the
2757 spa_base64_to_bits() function which I hope fixes it.
2760 36. The daemon start-up calls getloadavg() while still root for those OS that
2761 need the first call to be done as root, but it missed one case: when
2762 deliver_queue_load_max is set with deliver_drop_privilege. This is
2763 necessary for the benefit of the queue runner, because there is no re-exec
2764 when deliver_drop_privilege is set.
2766 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2767 was in use, was not putting the data itself into the right store pool;
2768 consequently, it could be overwritten for a subsequent message in the same
2769 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2772 38. Sometimes the final signoff response after QUIT could fail to get
2773 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2774 < 0 before doing a fflush(). This bug looks as though it goes back to the
2775 introduction of TLS in release 3.20, but "sometimes" must have been rare
2776 because the tests only now provoked it.
2782 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2783 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2784 Now it returns an integer. A number of calls were still expecting a T/F
2785 return. Fortuitously, in all cases, the tests worked in OK situations,
2786 which is the norm. However, things would have gone wrong on any write
2787 failures on the smtp file descriptor. This function is used when sending
2788 messages over SMTP and also when doing verify callouts.
2790 2. When Exim is called to do synchronous delivery of a locally submitted
2791 message (the -odf or -odi options), it no longer closes stderr before doing
2794 3. Implemented the mua_wrapper option.
2796 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2798 5. Implemented the functions header_remove(), header_testname(),
2799 header_add_at_position(), and receive_remove_recipient(), and exported them
2802 6. If an ACL "warn" statement specified the addition of headers, Exim already
2803 inserted X-ACL-Warn: at the start if there was no header name. However, it
2804 was not making this test for the second and subsequent header lines if
2805 there were newlines in the string. This meant that an invalid header could
2806 be inserted if Exim was badly configured.
2808 7. Allow an ACL "warn" statement to add header lines at the start or after all
2809 the Received: headers, as well as at the end.
2811 8. Added the rcpt_4xx retry error code.
2813 9. Added postmaster_mailfrom=xxx to callout verification option.
2815 10. Added mailfrom=xxxx to the callout verification option, for verify=
2818 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2819 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2821 12. Inserted some casts to stop certain compilers warning when using pointer
2822 differences as field lengths or precisions in printf-type calls (mostly
2823 affecting debugging statements).
2825 13. Added optional readline() support for -be (dynamically loaded).
2827 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2828 same clock tick as a message's arrival, so that its received time was the
2829 same as the "first fail" time on the retry record, and that message
2830 remained on the queue past the ultimate address timeout, every queue runner
2831 would try a delivery (because it was past the ultimate address timeout) but
2832 after another failure, the ultimate address timeout, which should have then
2833 bounced the address, did not kick in. This was a "< instead of <=" error;
2834 in most cases the first failure would have been in the next clock tick
2835 after the received time, and all would be well.
2837 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2838 being recognized when the domain list was tested by the match_domain
2839 condition in an expansion string.
2841 16. Added the ${str2b64: operator.
2843 17. Exim was always calling setrlimit() to set a large limit for the number of
2844 processes, without checking whether the existing limit was already
2845 adequate. (It did check for the limit on file descriptors.) Furthermore,
2846 errors from getrlimit() and setrlimit() were being ignored. Now they are
2847 logged to the main and panic logs, but Exim does carry on, to try to do its
2848 job under whatever limits there are.
2850 18. Imported PCRE 5.0.
2852 19. Trivial typo in log message " temporarily refused connection" (the leading
2855 20. If the log selector return_path_on_delivery was set and an address was
2856 redirected to /dev/null, the delivery process crashed because it assumed
2857 that a return path would always be set for a "successful" delivery. In this
2858 case, the whole delivery is bypassed as an optimization, and therefore no
2861 21. Internal re-arrangement: the function for sending a challenge and reading
2862 a response while authentication was assuming a zero-terminated challenge
2863 string. It's now changed to take a pointer and a length, to allow for
2864 binary data in such strings.
2866 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2868 23. Exim was not respecting finduser_retries when seeking the login of the
2869 uid under which it was called; it was always trying 10 times. (The default
2870 setting of finduser_retries is zero.) Also, it was sleeping after the final
2871 failure, which is pointless.
2873 24. Implemented tls_on_connect_ports.
2875 25. Implemented acl_smtp_predata.
2877 26. If the domain in control=submission is set empty, Exim assumes that the
2878 authenticated id is a complete email address when it generates From: or
2879 Sender: header lines.
2881 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2882 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2883 chown and chgrp in /bin and hostname in /usr/bin.
2885 28. Exim was keeping the "process log" file open after each use, just as it
2886 does for the main log. This opens the possibility of it remaining open for
2887 long periods when the USR1 signal hits a daemon. Occasional processlog
2888 errors were reported, that could have been caused by this. Anyway, it seems
2889 much more sensible not to leave this file open at all, so that is what now
2892 29. The long-running daemon process does not normally write to the log once it
2893 has entered its main loop, and it closes the log before doing so. This is
2894 so that log files can straightforwardly be renamed and moved. However,
2895 there are a couple of unusual error situations where the daemon does write
2896 log entries, and I had neglected to close the log afterwards.
2898 30. The text of an SMTP error response that was received during a remote
2899 delivery was being truncated at 512 bytes. This is too short for some of
2900 the long messages that one sometimes sees. I've increased the limit to
2903 31. It is now possible to make retry rules that apply only when a message has a
2904 specific sender, in particular, an empty sender.
2906 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2907 it possible to be selective about when SMTP synchronization is enforced.
2909 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2911 32. Implemented hosts_connection_nolog.
2913 33. Added an ACL for QUIT.
2915 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2918 35. Added mailbox_size and mailbox_filecount to appendfile.
2920 36. Added control = no_multiline_responses to ACLs.
2922 37. There was a bug in the logic of the code that waits for the clock to tick
2923 in the case where the clock went backwards by a substantial amount such
2924 that the microsecond fraction of "now" was more than the microsecond
2925 fraction of "then" (but the whole seconds number was less).
2927 38. Added support for the libradius Radius client library this is found on
2928 FreeBSD (previously only the radiusclient library was supported).
2934 1. When certain lookups returned multiple values in the form name=value, the
2935 quoting of the values was not always being done properly. Specifically:
2936 (a) If the value started with a double quote, but contained no whitespace,
2938 (b) If the value contained whitespace other than a space character (i.e.
2939 tabs or newlines or carriage returns) it was not quoted.
2940 This fix has been applied to the mysql and pgsql lookups by writing a
2941 separate quoting function and calling it from the lookup code. The fix
2942 should probably also be applied to nisplus, ibase and oracle lookups, but
2943 since I cannot test any of those, I have not disturbed their existing code.
2945 2. A hit in the callout cache for a specific address caused a log line with no
2946 reason for rejecting RCPT. Now it says "Previous (cached) callout
2947 verification failure".
2949 3. There was an off-by-one bug in the queryprogram router. An over-long
2950 return line was truncated at 256 instead of 255 characters, thereby
2951 overflowing its buffer with the terminating zero. As well as fixing this, I
2952 have increased the buffer size to 1024 (and made a note to document this).
2954 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2955 when Exim is waiting for an SMTP response from a remote server, Exim
2956 restarts its select() call on the socket, thereby resetting its timeout.
2957 This is not a problem when such interrupts are rare. Somebody set up a cron
2958 job to run exiwhat every 2 minutes, which is less than the normal select()
2959 timeout (5 or 10 minutes). This meant that the select() timeout never
2960 kicked in because it was always reset. I have fixed this by comparing the
2961 time when an interrupt arrives with the time at the start of the first call
2962 to select(). If more time than the timeout has elapsed, the interrupt is
2963 treated as a timeout.
2965 5. Some internal re-factoring in preparation for the addition of Sieve
2966 extensions (by MH). In particular, the "personal" test is moved to a
2967 separate function, and given an option for scanning Cc: and Bcc: (which is
2968 not set for Exim filters).
2970 6. When Exim created an email address using the login of the caller as the
2971 local part (e.g. when creating a From: or Sender: header line), it was not
2972 quoting the local part when it contained special characters such as @.
2974 7. Installed new OpenBSD configuration files.
2976 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2977 try to make them clearer.
2979 9. Callout options, other than the timeout value, were being ignored when
2980 verifying sender addresses in header lines. For example, when using
2982 verify = header_sender/callout=no_cache
2984 the cache was (incorrectly) being used.
2986 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2987 only the Cygwin environment.
2989 11. When return_path_on_delivery was set as a log selector, if different remote
2990 addresses in the same message used different return paths and parallel
2991 remote delivery occurred, the wrong values would sometimes be logged.
2992 (Whenever a remote delivery process finished, the return path value from
2993 the most recently started remote delivery process was logged.)
2995 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2996 lines when AUTH and/or TLS are in use. This is the "received protocol"
2997 field. Exim used to use "asmtp" for authenticated SMTP, without any
2998 indication (in the protocol name) for TLS use. Now it follows the RFC and
2999 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3000 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3001 names appear in log lines as well as in Received: header lines.
3003 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3004 extensions, and comparison tests, and to fix some bugs.
3006 14. Changes to the "personal" filter test:
3008 (1) The test was buggy in that it was just doing the equivalent of
3009 "contains" tests on header lines. For example, if a user's address was
3010 anne@some.where, the "personal" test would incorrectly be true for
3012 To: susanne@some.where
3014 This test is now done by extracting each address from the header in turn,
3015 and checking the entire address. Other tests that are part of "personal"
3016 are now done using regular expressions (for example, to check local parts
3017 of addresses in From: header lines).
3019 (2) The list of non-personal local parts in From: addresses has been
3020 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3021 taken from the Sieve specification recommendations.
3023 (3) If the message contains any header line starting with "List-" it is
3024 treated as non-personal.
3026 (4) The test for "circular" in the Subject: header line has been removed
3027 because it now seems ill-conceived.
3029 15. Minor typos in src/EDITME comments corrected.
3031 16. Installed latest exipick from John Jetmore.
3033 17. If headers_add on a router specified a text string that was too long for
3034 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3035 of string_sprintf() is now avoided.
3037 18. $message_body_size was not set (it was always zero) when running the DATA
3038 ACL and the local_scan() function.
3040 19. For the "mail" command in an Exim filter, no default was being set for
3041 the once_repeat time, causing a random time value to be used if "once" was
3042 specified. (If the value happened to be <= 0, no repeat happened.) The
3043 default is now 0s, meaning "never repeat". The "vacation" command was OK
3044 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3045 (I found it when inspecting the code).
3047 20. There is now an overall timeout for performing a callout verification. It
3048 defaults to 4 times the callout timeout, which applies to individual SMTP
3049 commands during the callout. The overall timeout applies when there is more
3050 than one host that can be tried. The timeout is checked before trying the
3051 next host. This prevents very long delays if there are a large number of
3052 hosts and all are timing out (e.g. when the network connections are timing
3053 out). The value of the overall timeout can be changed by specifying an
3054 additional sub-option for "callout", called "maxwait". For example:
3056 verify = sender/callout=5s,maxwait=20s
3058 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3059 to the end before writing, but this should make it even safer).
3061 22. Exim was forgetting that it had advertised PIPELINING for the second and
3062 subsequent messages on an SMTP connection. It was also not resetting its
3063 memory on STARTTLS and an internal HELO.
3065 23. When Exim logs an SMTP synchronization error within a session, it now
3066 records whether PIPELINING has been advertised or not.
3068 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3069 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3070 rather than long int.
3072 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3074 26. Added the never_mail option to autoreply.
3080 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3081 crash if the getsockname() call failed; this can happen if a connection is
3082 closed very soon after it is established. The problem was simply in the
3083 order in which certain operations were done, causing Exim to try to write
3084 to the SMTP stream before it had set up the file descriptor. The bug has
3085 been fixed by making things happen in the correct order.
3091 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3092 before the connection was closed, thus losing the rejection response.
3094 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3095 some early Solaris releases, but causes trouble in current releases where
3096 socklen_t is defined.
3098 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3101 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3104 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3105 that releasing the top of it at the end releases what was used for sub-
3106 expansions (unless the block got too big). However, discard this block if
3107 the first thing is a variable or header, so that we can use its block when
3108 it is dynamic (useful for very large $message_headers, for example).
3110 6. Lookups now cache *every* query, not just the most recent. A new, separate
3111 store pool is used for this. It can be recovered when all lookup caches are
3112 flushed. Lookups now release memory at the end of their result strings.
3113 This has involved some general refactoring of the lookup sources.
3115 7. Some code has been added to the store_xxx() functions to reduce the amount
3116 of flapping under certain conditions.
3118 8. log_incoming_interface used to affect only the <= reception log lines. Now
3119 it causes the local interface and port to be added to several more SMTP log
3120 lines, for example "SMTP connection from", and rejection lines.
3122 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3124 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3126 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3127 could be overwritten at the end of the current message (or the start of a
3128 new message if it was set in a HELO ACL). The value is now preserved for
3129 the duration of the SMTP connection.
3131 12. If a transport had a headers_rewrite setting, and a matching header line
3132 contained an unqualified address, that address was qualified, even if it
3133 did not match any rewriting rules. The underlying bug was that the values
3134 of the flags that permit the existence of unqualified sender and recipient
3135 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3136 non-local messages, and by -bnq for local messages) were not being
3137 preserved with the message after it was received.
3139 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3140 "next input=" as part of the text comprising the host identity instead of
3141 the correct text. The code was using the same buffer for two different
3142 strings. However, depending on which order the printing function evaluated
3143 its arguments, the bug did not always show up. Under Linux, for example, my
3144 test suite worked just fine.
3146 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3147 doesn't work with some older versions of Perl. It has been changed to "my",
3148 which in any case is probably the better facility to use.
3150 15. A really picky compiler found some instances of statements for creating
3151 error messages that either had too many or two few arguments for the format
3154 16. The size of the buffer for calls to the DNS resolver has been increased
3155 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3156 for addresses that have a lot of PTR records. This alleviates a problem; it
3157 does not fully solve it.
3159 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3160 buffer now truncates the list and logs the incident, which is the same
3161 action as happens when Exim is looking up a host name and its aliases.
3162 Previously in this situation something unpredictable would happen;
3163 sometimes it was "internal error: store_reset failed".
3165 18. If a server dropped the connection unexpectedly when an Exim client was
3166 using GnuTLS and trying to read a response, the client delivery process
3167 crashed while trying to generate an error log message.
3169 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3170 string, for example:
3172 warn message = H1: something\nH2: something
3174 the text was added as a single header line from Exim's point of view
3175 though it ended up OK in the delivered message. However, searching for the
3176 second and subsequent header lines using $h_h2: did not work. This has been
3177 fixed. Similarly, if a system filter added multiple headers in this way,
3178 the routers could not see them.
3180 20. Expanded the error message when iplsearch is called with an invalid key to
3181 suggest using net-iplsearch in a host list.
3183 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3184 ACLs are no longer actually imposed (and a message to that effect is
3187 22. If a "gecos" field in a passwd entry contained escaped characters, in
3188 particular, if it contained a \" sequence, Exim got it wrong when building
3189 a From: or a Sender: header from that name. A second bug also caused
3190 incorrect handling when an unquoted " was present following a character
3191 that needed quoting.
3193 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3194 was not being matched caselessly.
3196 24. Arranged for all hyphens in the exim.8 source to be escaped with
3199 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3200 properly. Recipient callout cache records were still being keyed to include
3201 the sender, even when use_sender was set false. This led to far more
3202 callouts that were necessary. The sender is no longer included in the key
3203 when use_sender is false.
3205 26. Added "control = submission" modifier to ACLs.
3207 27. Added the ${base62d: operator to decode base 62 numbers.
3209 28. dnsdb lookups can now access SRV records.
3211 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3212 the configuration file.
3214 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3215 (-v) mode. This makes the output for a verbose queue run more intelligible.
3217 31. Added a use_postmaster feature to recipient callouts.
3219 32. Added the $body_zerocount variable, containing the number of binary zero
3220 bytes in the message body.
3222 33. The time of last modification of the "new" subdirectory is now used as the
3223 "mailbox time last read" when there is a quota error for a maildir
3226 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3228 35. Added +ignore_unknown as a special item in host lists.
3230 36. Code for decoding IPv6 addresses in host lists is now included, even if
3231 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3232 address was recognized as an IP address, but was then not correctly decoded
3233 into binary, causing unexpected and incorrect effects when compared with
3240 1. Very minor rewording of debugging text in manualroute to say "list of
3241 hosts" instead of "hostlist".
3243 2. If verify=header_syntax was set, and a header line with an unqualified
3244 address (no domain) and a large number of spaces between the end of the
3245 name and the colon was received, the reception process suffered a buffer
3246 overflow, and (when I tested it) crashed. This was caused by some obsolete
3247 code that should have been removed. The fix is to remove it!
3249 3. When running in the test harness, delay a bit after writing a bounce
3250 message to get a bit more predictability in the log output.
3252 4. Added a call to search_tidyup() just before forking a reception process. In
3253 theory, someone could use a lookup in the expansion of smtp_accept_max_
3254 per_host which, without the tidyup, could leave open a database connection.
3256 5. Added the variables $recipient_data and $sender_data which get set from a
3257 lookup success in an ACL "recipients" or "senders" condition, or a router
3258 "senders" option, similar to $domain_data and $local_part_data.
3260 6. Moved the writing of debug_print from before to after the "senders" test
3263 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3264 problems for message scanning, either using a data ACL, or using
3265 local_scan() because the Received: header was not generated till after they
3266 were called (in order to set the time as the time of reception completion).
3267 I have revised the way this works. The header is now generated after the
3268 body is received, but before the ACL or local_scan() are called. After they
3269 are run, the timestamp in the header is updated.
3275 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3276 before starting a queue runner without re-exec. This happened only when
3277 deliver_drop_privilege was set or when the Exim user was set to root. The
3278 effect of the bug was that timeouts during subsequent deliveries caused
3279 crashes instead of being properly handled. The handler is now left at its
3280 default (and expected) setting.
3282 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3283 message, again when deliver_drop_privilege is set or Exim is run as root.
3284 The bug described in (1) was not present in this case, but the tidying up
3285 of the other signals was missing. I have made the two cases consistent.
3287 3. The ignore_target_hosts setting on a manualroute router was being ignored
3288 for hosts that were looked up using the /MX notation.
3290 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3293 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3294 operated on the sender address. After changing the $sender_address to <>
3295 for the sender address verify, Exim was re-instated it as the original
3296 (before rewriting) address, but remembering that it had rewritten it, so it
3297 wasn't rewriting it again. This bug also had the effect of breaking the
3298 sender address verification caching when the sender address was rewritten.
3300 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3301 This has been changed so that if the ip literal address matches
3302 ignore_target_hosts, the router declines.
3304 7. Added expansion conditions match_domain, match_address, and match_local_
3305 part (NOT match_host).
3307 8. The placeholder for the Received: header didn't have a length field set.
3309 9. Added code to Exim itself and to exim_lock to test for a specific race
3310 condition that could lead to file corruption when using MBX delivery. The
3311 issue is with the lockfile that is created in /tmp. If this file is removed
3312 after a process has opened it but before that process has acquired a lock,
3313 there is the potential for a second process to recreate the file and also
3314 acquire a lock. This could lead to two Exim processes writing to the file
3315 at the same time. The added code performs the same test as UW imapd; it
3316 checks after acquiring the lock that its file descriptor still refers to
3317 the same named file.
3319 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3320 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3323 11. Added the smtp_active_hostname option. If used, this will typically be made
3324 to depend on the incoming interface address. Because $interface_address is
3325 not set up until the daemon has forked a reception process, error responses
3326 that can happen earlier (such as "too many connections") no longer contain
3329 12. If an expansion in a condition on a "warn" statement fails because a lookup
3330 defers, the "warn" statement is abandoned, and the next ACL statement is
3331 processed. Previously this caused the whole ACL to be aborted.
3333 13. Added the iplsearch lookup type.
3335 14. Added ident_timeout as a log selector.
3337 15. Added tls_certificate_verified as a log selector.
3339 16. Added a global option tls_require_ciphers (compare the smtp transport
3340 option of the same name). This controls incoming TLS connections.
3342 17. I finally figured out how to make tls_require_ciphers do a similar thing
3343 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3344 before starting the TLS session.
3346 18. Tabs are now shown as \t in -bP output.
3348 19. If the log selector return_path_on_delivery was set, Exim crashed when
3349 bouncing a message because it had too many Received: header lines.
3351 20. If two routers both had headers_remove settings, and the first one included
3352 a superfluous trailing colon, the final name in the first list and the
3353 first name in the second list were incorrectly joined into one item (with a
3354 colon in the middle).
3360 1. Added -C and -D options to the exinext utility, mainly to make it easier
3361 to include in the automated testing, but these could be helpful when
3362 multiple configurations are in use.
3364 2. The exinext utility was not formatting the output nicely when there was
3365 an alternate port involved in the retry record key, nor when there was a
3366 message id as well (for retries that were specific to a specific message
3367 and a specific host). It was also confused by IPv6 addresses, because of
3368 the additional colons they contain. I have fixed the IPv4 problem, and
3369 patched it up to do a reasonable job for IPv6.
3371 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3372 delivery, the log line now contains "pipelined" if PIPELINING was used.
3374 4. An SMTP transport process used to panic and die if the bind() call to set
3375 an explicit outgoing interface failed. This has been changed; it is now
3376 treated in the same way as a connect() failure.
3378 5. A reference to $sender_host_name in the part of a conditional expansion
3379 that was being skipped was still causing a DNS lookup. This no longer
3382 6. The def: expansion condition was not recognizing references to header lines
3383 that used bh_ and bheader_.
3385 7. Added the _cache feature to named lists.
3387 8. The code for checking quota_filecount in the appendfile transport was
3388 allowing one more file than it should have been.
3390 9. For compatibility with Sendmail, the command line option
3398 and sets the incoming protocol and host name (for trusted callers). The
3399 host name and its colon can be omitted when only the protocol is to be set.
3400 Note the Exim already has two private options, -pd and -ps, that refer to
3401 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3402 "s", but I don't think that's a major issue.
3404 10. A number of refactoring changes to the code, none of which should affect
3407 (a) The number of logging options was getting close to filling up the
3408 32-bit word that was used as a bit map. I have split them into two classes:
3409 those that are passed in the argument to log_write(), and those that are
3410 only ever tested independently outside of that function. These are now in
3411 separate 32-bit words, so there is plenty of room for expansion again.
3412 There is no change in the user interface or the logging behaviour.
3414 (b) When building, for example, log lines, the code previously used a
3415 macro that called string_cat() twice, in order to add two strings. This is
3416 not really sufficiently general. Furthermore, there was one instance where
3417 it was actually wrong because one of the argument was used twice, and in
3418 one call a function was used. (As it happened, calling the function twice
3419 did not affect the overall behaviour.) The macro has been replaced by a
3420 function that can join an arbitrary number of extra strings onto a growing
3423 (c) The code for expansion conditions now uses a table and a binary chop
3424 instead of a serial search (which was left over from when there were very
3425 few conditions). Also, it now recognizes conditions like "pam" even when
3426 the relevant support is not compiled in: a suitably worded error message is
3427 given if an attempt is made to use such a condition.
3429 11. Added ${time_interval:xxxxx}.
3431 12. A bug was causing one of the ddress fields not to be passed back correctly
3432 from remote delivery subprocesses. The field in question was not being
3433 subsequently used, so this caused to problems in practice.
3435 13. Added new log selectors queue_time and deliver_time.
3437 14. Might have fixed a bug in maildirsizefile handling that threw up
3438 "unexpected character" debug warnings, and recalculated the data
3439 unnecessarily. In any case, I expanded the warning message to give more
3442 15. Added the message "Restricted characters in address" to the statements in
3443 the default ACL that block characters like @ and % in local parts.
3445 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3446 Three changes have been made:
3448 (a) There was a serious bug; a negative response to MAIL caused the whole
3449 recipient domain to be cached as invalid, thereby blocking all messages
3450 to all local parts at the same domain, from all senders. This bug has
3451 been fixed. The domain is no longer cached after a negative response to
3452 MAIL if the sender used is not empty.
3454 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3457 (c) A new callout option, "use_sender" has been added for people who want
3458 the modified behaviour.
3464 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3467 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3468 indeed breaks things for older releases.
3470 3. Added additional logging to the case where there is a problem reading data
3471 from a filter that is running in a subprocess using a pipe, in order to
3472 try to track down a specific problem.
3474 4. Testing facility fudge: when running in the test harness and attempting
3475 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3476 getting "No route to host". Convert this to a timeout.
3478 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3481 6. Some OS don't have socklen_t but use size_t instead. This affects the
3482 fifth argument of getsockopt() amongst other things. This is now
3483 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3484 can be set for individual OS. I have set it for SunOS5, OSF1, and
3485 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3486 some earlier ones do not.
3488 7. Change 4.30/15 was not doing the test caselessly.
3490 8. The standard form for an IPv6 address literal was being rejected by address
3491 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3492 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3493 this, as well as the form without the "IPv6" on the front (but only when
3494 address literals are enabled, of course).
3496 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3498 10. Exim crashed if a message with an empty sender address specified by -f
3499 encountered a router with an errors_to setting. This could be provoked only
3500 by a command such as
3504 where an empty string was supplied; "<>" did not hit this bug.
3506 11. Installed PCRE release 4.5.
3508 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3509 remained set. It is now erased.
3511 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3512 times from message ids (which are base 36 rather than the normal 62).
3514 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3515 were being counted as actual protocol errors, and logged if the log
3516 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3517 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3518 and DATA following a set of rejected RCPTs do not count as protocol errors.
3519 In other words, Exim assumes they were pipelined, though this may not
3520 actually be the case. Of course, in all cases the client gets an
3521 appropriate error code.
3523 15. If a lookup fails in an ACL condition, a message about the failure may
3524 be available; it is used if testing the ACL cannot continue, because most
3525 such messages specify what the cause of the deferral is. However, some
3526 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3527 that caused an old message to be retained and used if a later statement
3528 caused a defer, replacing the real cause of the deferral.
3530 16. If an IP address had so many PTR records that the DNS lookup buffer
3531 was not large enough to hold them, Exim could crash while trying to process
3532 the truncated data. It now detects and logs this case.
3534 17. Further to 4.21/58, another change has been made: if (and only if) the
3535 first line of a message (the first header line) ends with CRLF, a bare LF
3536 in a subsequent header line has a space inserted after it, so as not to
3537 terminate the header.
3539 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3540 unnecessarily, used atoi() instead of strtol(), and didn't check the
3541 termination when getting file sizes from file names by regex.
3543 19. Completely re-implemented the support for maildirsize files, in the light
3544 of a number of problems with the previous contributed implementation
3545 (4.30/29). In particular:
3547 . If the quota is zero, the maildirsize file is maintained, but no quota is
3550 . If the maildir directory does not exist, it is created before any attempt
3551 to write a maildirsize file.
3553 . The quota value in the file is just a cache; if the quota is changed in
3554 the transport, the new value overrides.
3556 . A regular expression is available for excluding directories from the
3559 20. The autoreply transport checks the characters in options that define the
3560 message's headers; it allows continued headers, but it was checking with
3561 isspace() after an embedded newline instead of explicitly looking for a
3564 21. If all the "regular" hosts to which an address was routed had passed their
3565 expiry times, and had not reached their retry times, the address was
3566 bounced, even if fallback hosts were defined. Now Exim should go on to try
3569 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3570 equivalent code in the SMTP transport. Some hosts send humungous responses
3571 to HELO/EHLO, more than 1024 it seems.
3573 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3574 gives compiler warnings in some environments. I've now done it "properly",
3577 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3578 (because of problems with the built-in one) was declared to return uschar *
3579 instead of char *, causing compiler failure.
3581 25. Fixed a file descriptor leak when processing alias/forward files.
3583 26. Fixed a minor format string issue in dbfn.c.
3585 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3587 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3588 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3591 29. When a sender address is verified, it is cached, to save repeating the test
3592 when there is more than one recipient in a message. However, when the
3593 verification involves a callout, it is possible for different callout
3594 options to be set for different recipients. It is too complicated to keep
3595 track of this in the cache, so now Exim always runs a verification when a
3596 callout is required, relying on the callout cache for the optimization.
3597 The overhead is duplication of the address routing, but this should not be
3600 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3601 to be verified with callout=postmaster, and the main callout worked but the
3602 postmaster check failed, the verification correctly failed. However, if a
3603 subsequent RCPT command asked for sender verification *without* the
3604 postmaster check, incorrect caching caused this verification also to fail,
3607 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3608 it was not caching the DNS options (qualify_single, search_parents) that
3609 were used when the lookup failed. A subsequent lookup with different
3610 options therefore always gave the same answer, though there were cases
3611 where it should not have. (Example: a "domains = !$mx_any" option on a
3612 dnslookup router: the "domains" option is always processed without any
3613 widening, but the router might have qualify_single set.) Now Exim uses the
3614 cached value only when the same options are set.
3616 32. Added John Jetmore's "exipick" utility to the distribution.
3618 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3619 than a timeout (e.g. a certificate is required, and is not provided), an
3620 Exim server now closes the connection immediately. Previously it waited for
3621 the client to close - but if the client is SSL, it seems that they each
3622 wait for each other, leading to a delay before one of them times out.
3624 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3625 maintained 0.8.x compatibility because I don't think many are using it, and
3626 it is clearly obsolete.
3628 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3631 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3632 client certificate was expired. A simple patch fixes this, though I don't
3633 understand the full logic of why the verify callback is called multiple
3636 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3637 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3638 which causes problems with some clients (such as the Certicom SSL Plus
3639 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3640 disables the coutermeasure allowing Eudora to connect."
3642 38. Exim was not checking that a write() to a log file succeeded. This could
3643 lead to Bad Things if a log got too big, in particular if it hit a file
3644 size limit. Exim now panics and dies if it cannot write to a log file, just
3645 as it does if it cannot open a log file.
3647 39. Modified OS/Makefile-Linux so that it now contains
3649 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3651 The two -D definitions ensure that Exim is compiled with large file
3652 support, which makes it possible to handle log files that are bigger than
3655 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3656 instance) a domain was checked against a named list that involved a lookup,
3657 causing $domain_data to be set, then another domain was checked against the
3658 same list, then the first domain was re-checked, the value of $domain_data
3659 after the final check could be wrong. In particular, if the second check
3660 failed, it could be set empty. This bug probably also applied to
3663 41. The strip_trailing_dot option was not being applied to the address given
3664 with the -f command-line option.
3666 42. The code for reading a message's header from the spool was incrementing
3667 $received_count, but never initializing it. This meant that the value was
3668 incorrect (doubled) while delivering a message in the same process in which
3669 it was received. In the most common configuration of Exim, this never
3670 happens - a fresh exec is done - but it can happen when
3671 deliver_drop_privilege is set.
3673 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3674 it now includes up to 150 characters of the unexpected data in the log
3677 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3678 and building data strings. The size of both of these buffers was 10 000
3679 bytes - far larger than anybody would *ever* want, thought I. Needless to
3680 say, somebody hit the limit. I have increased the maximum line length to
3681 20 000 and the maximum data length of concatenated lines to 100 000. I have
3682 also fixed two bugs, because there was no checking on these buffers. Tsk,
3683 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3684 buffer is too small.
3686 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3687 lsearch lookups. Now it does.
3689 46. When parsing a route_list item in a manualroute router, a fixed-length
3690 buffer was used for the list of hosts. I made this 1024 bytes long,
3691 thinking that nobody would ever have a list of hosts that long. Wrong.
3692 Somebody had a whole pile of complicated expansion conditions, and the
3693 string was silently truncated, leading to an expansion error. It turns out
3694 that it is easier to change to an unlimited length (owing to other changes
3695 that have happened since this code was originally written) than to build
3696 structure for giving a limitation error. The length of the item that
3697 expands into the list of hosts is now unlimited.
3699 47. The lsearch lookup could not handle data where the length of text line was
3700 more than 4095 characters. Such lines were truncated, leading to shortened
3701 data being returned. It should now handle lines of any length.
3703 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3704 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3707 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3708 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3709 no longer happens in comment lines. A list of replacements is now placed
3710 at the head of all of the source files, except those whose only change is
3711 to replace PERL_COMMAND in the very first #! line.
3713 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3714 messages on the queue, with a bottom-up merge sort, using code contributed
3715 by Michael Haardt. This should make operations like -bp somewhat faster on
3716 large queues. It won't affect queue runners, except when queue_run_in_order
3719 51. Installed eximstats 1.31 in the distribution.
3721 52. Added support for SRV lookups to the dnslookup router.
3723 53. If an ACL referred to $message_body or $message_body_end, the value was not
3724 reset for any messages that followed in the same SMTP session.
3726 54. The store-handling optimization for building very long strings was not
3727 differentiating between the different store pools. I don't think this
3728 actually made any difference in practice, but I've tidied it.
3730 55. While running the routers to verify a sender address, $sender_address
3731 was still set to the sender address. This is wrong, because when routing to
3732 send a bounce to the sender, it would be empty. Therefore, I have changed
3733 it so that, while verifying a sender address, $sender_address is set to <>.
3734 (There is no change to what happens when verifying a recipient address.)
3736 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3737 target A or AAAA records (if not already returned) without resetting the
3738 qualify_single or search_parents options of the DNS resolver. These are
3739 inappropriate in this case because the targets of MX and SRV records must
3740 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3741 target that, when qualified, matched something in the local domain. These
3742 two options are now turned off when doing these lookups.
3744 57. It seems that at least some releases of Reiserfs (which does not have the
3745 concept of a fixed number of inodes) returns zero and not -1 for the
3746 number of available inodes. This interacted badly with check_spool_inodes,
3747 which assumed that -1 was the "no such thing" setting. What I have done is
3748 to check that the total number of inodes is greater than zero before doing
3749 the test of how many are available.
3751 58. When a "warn" ACL statement has a log_message modifier, the message is
3752 remembered, and not repeated. This is to avoid a lot of repetition when a
3753 message has many recipients that cause the same warning to be written.
3754 Howewer, Exim was preserving the list of already written lines for an
3755 entire SMTP session, which doesn't seem right. The memory is now reset if a
3756 new message is started.
3758 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3759 debugging output unless log_rewrite was also set.
3761 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3762 of (int)(handle) when we know that handle contains (void *)(-1).
3764 61. The Exim daemon panic-logs an error return when it closes the incoming
3765 connection. However "connection reset by peer" seems to be common, and
3766 isn't really an error worthy of noting specially, so that particular error
3769 62. When Exim is trying to find all the local interfaces, it used to panic and
3770 die if the ioctl to get the interface flags failed. However, it seems that
3771 on at least one OS (Solaris 9) it is possible to have an interface that is
3772 included in the list of interfaces, but for which you get a failure error
3773 for this call. This happens when the interface is not "plumbed" into a
3774 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3775 failure of the "get flags" call assumes that the interface is down.
3777 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3778 makes life easier for people who are doing arithmetic on fields extracted
3779 from dates, where you often get leading zeros that should not be
3780 interpreted as octal.
3782 64. Added qualify_domain to the redirect router, to override the global
3785 65. If a pathologically long header line contained very many addresses (the
3786 report of this problem mentioned 10 000) and each of them was rewritten,
3787 Exim could use up a very large amount of memory. (It kept on making new
3788 copies of the header line as it rewrote, and never released the old ones.)
3789 At the expense of a bit more processing, the header rewriting function has
3790 been changed so that it no longer eats memory in this way.
3792 66. The generation of the Received: header has been moved from the time that a
3793 message starts to be received, to the time that it finishes. The timestamp
3794 in the Received: header should now be very close to that of the <= log
3795 line. There are two side-effects of this change:
3797 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3798 logged header lines no longer include the local Received: line, because
3799 it has not yet been created. The same applies to a copy of the message
3800 that is returned to a non-SMTP sender when a message is rejected.
3802 (b) When a filter file is tested using -bf, no additional Received: header
3803 is added to the test message. After some thought, I decided that this
3806 This change does not affect the value of $received_for. It is still set
3807 after address rewriting, but before local_scan() is called.
3809 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3811 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3812 gave an unhelpful panic error message, and a defer error. I have managed to
3813 change this behaviour so that it now rejects any supplied certificate,
3814 which seems right, as the list of acceptable certificates is empty.
3816 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3817 gave an unhelpful defer error. I have not managed to make this reject any
3818 supplied certificates, but the error message it gives is "no certificate
3819 supplied", which is not helpful.
3821 70. exigrep's output now also includes lines that are not associated with any
3822 message, but which match the given pattern. Implemented by a patch from
3823 Martin Sluka, which also tidied up the Perl a bit.
3825 71. Recipient callout verification, like sender verification, was using <> in
3826 the MAIL FROM command. This isn't really the right thing, since the actual
3827 sender may affect whether the remote host accepts the recipient or not. I
3828 have changed it to use the actual sender in the callout; this means that
3829 the cache record is now keyed on a recipient/sender pair, not just the
3830 recipient address. There doesn't seem to be a real danger of callout loops,
3831 since a callout by the remote host to check the sender would use <>.
3832 [SEE ABOVE: changed after hitting problems.]
3834 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3835 temporary errors. However, in the case of such a code being given after
3836 the end of a data transmission (i.e. after ".") Exim was failing to write
3837 a retry record for the message. (Yes, there was some broken host that was
3838 actually sending 8xx at this point.)
3840 73. An unknown lookup type in a host list could cause Exim to panic-die when
3841 the list was checked. (An example that provoked this was putting <; in the
3842 middle of a list instead of at the start.) If this happened during a DATA
3843 ACL check, a -D file could be left lying around. This kind of configuration
3844 error no longer causes Exim to die; instead it causes a defer errror. The
3845 incident is still logged to the main and panic logs.
3847 74. Buglet left over from Exim 3 conversion. The message "too many messages
3848 in one connection" was written to the rejectlog but not the mainlog, except
3849 when address rewriting (yes!) was being logged.
3851 75. Added write_rejectlog option.
3853 76. When a system filter was run not as root (that is, when system_filter_user
3854 was set), the values of the $n variables were not being returned to the
3855 main process; thus, they were not subsequently available in the $sn
3858 77. Added +return_path_on_delivery log selector.
3860 78. A connection timeout was being treated differently from recipients deferred
3861 when testing hosts_max_try with a message that was older than the host's
3862 retry timeout. (The host should not be counted, thus allowing all hosts to
3863 be tried at least once before bouncing.) This may have been the cause of an
3864 occasionally reported bug whereby a message would remain on the queue
3865 longer than the retry timeout, but would be bounced if a delivery was
3866 forced. I say "may" because I never totally pinned down the problem;
3867 setting up timeout/retry tests is difficult. See also the next item.
3869 79. The ultimate address timeout was not being applied to errors that involved
3870 a combination of host plus message (for example, a timeout on a MAIL
3871 command). When an address resolved to a number of possible hosts, and they
3872 were not all tried for each delivery (e.g. because of hosts_max_try), a
3873 message could remain on the queue longer than the retry timeout.
3875 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3878 81. Fixed an obscure SMTP outgoing bug which required at least the following
3879 conditions: (a) there was another message waiting for the same server;
3880 (b) the server returned 5xx to all RCPT commands in the first message so
3881 that the message was not completed; (c) the server dropped the connection
3882 or gave a negative response to the RSET that Exim sends to abort the
3883 transaction. The observed case was a dropped connection after DATA that had
3884 been sent in pipelining mode. That is, the server had advertised PIPELINING
3885 but was not implementing it correctly. The effect of the bug was incorrect
3886 behaviour, such as trying another host, and this could lead to a crash.
3892 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3893 and daemon.c were passed as pointers to ints; they should have been
3894 pointers to socklen_t variables (which are typically unsigned ints).
3896 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3899 3. Fixed a really odd bug that affected only the testing scheme; patching a
3900 certain fixed string in the binary changed the value of another string that
3901 happened to be identical to the end of the original first string.
3903 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3904 name", it returns that address as the IP address. On some operating
3905 systems (e.g. Solaris), it also passes back the IP address string as the
3906 "host name". However, on others (e.g. Linux), it passes back an empty
3907 string. Exim wasn't checking for this, and was changing the host name to an
3908 empty string, assuming it had been canonicized.
3910 5. Although rare, it is permitted to have more than one PTR record for a given
3911 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3912 all the names associated with an address, because they do in Solaris.
3913 However, it seems that they do not in Linux for data that comes from the
3914 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3915 I found this out when I moved to a new Linux workstation and tried to run
3916 the Exim test suite.
3918 To get round this problem I have changed the code so that it now does its
3919 own call to the DNS to look up PTR records when searching for a host name.
3920 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3921 addresses that are only in /etc/hosts are still found.
3923 This behaviour is, however, controlled by an option called host_lookup_
3924 order, which defaults to "bydns:byaddr". If people want to use the other
3925 order, or indeed, just use one or the other means of lookup, they can
3926 specify it in this variable.
3928 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3929 some operating systems, this comes back from gethostbyaddr() as an empty
3930 string, and this is what Exim used to test for. However, it seems that in
3931 other systems, "." is yielded. Exim now tests for this case too.
3933 7. The values of check_spool_space and check_log_space are now held internally
3934 as a number of kilobytes instead of an absolute number of bytes. If a
3935 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3936 kilobyte. This means that much larger values can be stored.
3938 8. Exim monitor: an attempt to get the action menu when not actually pointing
3939 at a message produces an empty menu entitled "No message selected". This
3940 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3941 no entries in it ("Shell widget menu has zero width and/or height"). So I
3942 have added a single, blank menu entry in this case.
3944 9. Added ${quote_local_part.
3946 10. MIME decoding is now applied to the contents of Subject: header lines when
3949 11. Now that a reference to $sender_host_address automatically causes a reverse
3950 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3951 host lookup before query-style lookups in lists that might use this
3952 variable. This has therefore been abolished, and the "net-" prefix is no
3953 longer necessary for query-style lookups.
3955 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3956 have been a typo for LFLAGS, so it has been changed.
3958 13. The install script calls Exim with "-C /dev/null" in order to find the
3959 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3960 to be output. Howeve, since Exim outputs its version number before the
3961 error, it didn't break the script. It just looked ugly. I fixed this by
3962 always allowing "-C /dev/null" if the caller is root.
3964 14. Ignore overlarge ACL variable number when reading spool file - insurance
3965 against a later release with more variables having written the file.
3967 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3968 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3969 this, as well as the form without the "IPv6" on the front.
3971 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3972 OS/Makefile-Darwin file.
3974 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3975 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3977 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3978 variable. However, some OS use ints for inodes. I've added cast to long int
3979 to get rid of the compiler warning.
3981 19. I had forgotten to lock out "/../" in configuration file names when
3982 ALT_CONFIG_PREFIX was set.
3984 20. Routers used for verification do not need to specify transports. However,
3985 if such a router generated a host list, and callout was configured, Exim
3986 crashed, because it could not find a port number from the (non-existent)
3987 transport. It now assumes port 25 in this circumstance.
3989 21. Added the -t option to exigrep.
3991 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3992 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3993 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3994 them is set, LOOKUP_LSEARCH is forced.
3996 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3998 24. Added sender and host information to the "rejected by local_scan()" log
3999 line; previously there was no indication of these.
4001 25. Added .include_if_exists.
4003 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4004 Linux. It turns out that not all file systems support this. Apparently some
4005 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4006 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4007 not supported on the file descriptor, is now ignored when Exim is trying to
4008 sync a directory. This applies only to Linux.
4010 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4012 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4013 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4016 29. Added support for maildirsize files from supplied patch (modified a bit).
4018 30. The use of :fail: followed by an empty string could lead Exim to respond to
4019 sender verification failures with (e.g.):
4021 550 Verification failed for <xxx>
4022 550 Sender verify failed
4024 where the first response line was missing the '-' that indicates it is not
4025 the final line of the response.
4027 31. The loop for finding the name of the user that called Exim had a hardwired
4028 limit of 10; it now uses the value of finduser_retries, which is used for
4029 all other user lookups.
4031 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4034 33. Exim was neglecting to zero errno before one call of strtol() when
4035 expanding a string and expecting an integer value. On some systems this
4036 resulted in spurious "integer overflow" errors. Also, it was casting the
4037 result into an int without checking.
4039 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4040 did not work. The code looks as if it has *never* worked, though it appears
4041 to have been documented since at least releast 1.62. I have made it work.
4043 35. The "timeout_DNS" error in retry rules, also documented since at least
4044 1.62, also never worked. As it isn't clear exactly what this means, and
4045 clearly it isn't a major issue, I have abolished the feature by treating it
4046 as "timeout", and writing a warning to the main and panic logs.
4048 36. The display of retry rules for -brt wasn't always showing the error code
4051 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4052 timeout_connect_A, timeout_connect_MX.
4054 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4055 to the empty sender.
4057 39. The daemon was not analysing the content of -oX till after it had closed
4058 stderr and disconnected from the controlling terminal. This meant that any
4059 syntax errors were only noted on the panic log, and the return code from
4060 the command was 0. By re-arranging the code a little, I've made the
4061 decoding happen first, so such errors now appear on stderr, and the return
4062 code is 1. However, the actual setting up of the sockets still happens in
4063 the disconnected process, so errors there are still only recorded on the
4066 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4067 connections (as happens on some IP stacks) was logged at start up time as
4068 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4069 it from "IPv6 and IPv4", which means that two separate sockets are being
4072 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4073 says whether AF_INET or AF_INET6 was passed as an argument.
4075 42. Exiwhat output was messed up when time zones were included in log
4078 43. Exiwhat now gives more information about the daemon's listening ports,
4079 and whether -tls-on-connect was used.
4081 44. The "port" option of the smtp transport is now expanded.
4083 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4084 silently ignored. Now an error message is written to the main and panic
4087 46. There's a new ACL modifier called "logwrite" which writes to a log file
4088 as soon as it is encountered.
4090 47. Added $local_user_uid and $local_user_gid at routing time.
4092 48. Exim crashed when trying to verify a sender address that was being
4095 49. Exim was recognizing only a space character after ".include". It now also
4096 recognizes a tab character.
4098 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4099 extracting the relevant information from the specification. The man page no
4100 longer contains scrambled data for the -d option, and I've added a section
4101 at the front about calling Exim under different names.
4103 51. Added "extra_headers" argument to the "mail" command in filter files.
4105 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4108 53. Installed eximstats 1.29.
4110 54. Added transport_filter_timeout as a generic transport option.
4112 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4113 Cc: header lines. This was required by RFC 822, but it not required by RFC
4116 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4117 incoming messages that did not have them. Now it does so only if the
4118 message originates locally, that is, if there is no associated remote host
4119 address. When Resent- header lines are present, this applies to the Resent-
4120 lines rather than the non-Resent- lines.
4122 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4123 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4125 58. Messages for configuration errors now include the name of the main
4126 configuration file - useful now that there may be more than one file in a
4127 list (.included file names were always shown).
4129 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4130 for those rare installations that do not start the daemon as root or run it
4131 setuid root. I've cut out the call to initgroups() if the daemon is not
4134 60. The Exim user and group can now be bound into the binary as text strings
4135 that are looked up at the start of Exim's processing.
4137 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4139 62. Added $mailstore_basename variable.
4141 63. Installed patch to sieve.c from Michael Haardt.
4143 64. When Exim failed to open the panic log after failing to open the main log,
4144 the original message it was trying to log was written to stderr and debug
4145 output, but if they were not available (the usual case in production), it
4146 was lost. Now it is written to syslog before the two lines that record the
4147 failures to open the logs.
4149 65. Users' Exim filters run in subprocesses under the user's uid. It is
4150 possible for a "deliver" command or an alias in a "personal" command to
4151 provoke an address rewrite. If logging of address rewriting is configured,
4152 this fails because the process is not running as root or exim. There may be
4153 a better way of dealing with this, but for the moment (because 4.30 needs
4154 to be released), I have disabled address rewrite logging when running a
4155 filter in a non-root, non-exim process.
4161 1. The buildconfig auxiliary program wasn't quoting the value set for
4162 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4163 not defined. This bug was masked in 4.22 by the effect that was fixed in
4166 2. Some messages that were rejected after a message id was allocated were
4167 shown as "incomplete" by exigrep. It no longer does this for messages that
4168 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4170 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4171 have allow_domain_literals set, the ID did not get logged in the <= line.
4172 Domain literals are now always recognized in Message-ID: header lines.
4174 4. The first argument for a ${extract expansion item is the key name or field
4175 number. Leading and trailing spaces in this item were not being ignored,
4176 causing some misleading effects.
4178 5. When deliver_drop_privilege was set, single queue runner processes started
4179 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4180 same command in the process it spins off) were not dropping privilege.
4182 6. When the daemon running as "exim" started a queue runner, it always
4183 re-executed Exim in the spun-off process. This is a waste of effort when
4184 deliver_drop_privilege is set. The new process now just calls the
4185 queue-runner function directly.
4191 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4194 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4195 ignored. Though the use of -oP was forcing the writing of a pid file, it
4196 was always written to the default place.
4198 3. If the message "no IP address found for host xxxx" is generated during
4199 incoming verification, it is now followed by identification of the incoming
4200 connection (so you can more easily find what provoked it).
4202 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4204 5. Added some features to "harden" Exim a bit more against certain attacks:
4206 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4207 be put in Local/Makefile. This is like the never_users runtime option,
4208 but it cannot be overridden. The default setting is "root".
4210 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4211 prefix string with which any file named in a -C command line option
4214 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4215 is retained for -C and -D only if the caller of Exim is root. Without
4216 it, the exim user may also use -C and -D and retain privilege.
4218 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4219 command line option is disabled.
4221 6. Macro names set by the -D option must start with an upper case letter, just
4222 like macro names defined in the configuration file.
4224 7. Added "dereference=" facility to LDAP.
4226 8. Two instances of the typo "uknown" in the source files are fixed.
4228 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4229 the Configure-Makefile script screwed up while processing it.
4231 10. Incorporated PCRE 4.4.
4233 11. The SMTP synchronization check was not operating right at the start of an
4234 SMTP session. For example, it could not catch a HELO sent before the client
4235 waited for the greeting. There is now a check for outstanding input at the
4236 point when the greeting is written. Because of the duplex, asynchronous
4237 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4238 way, but not yet received, when the check is performed.
4240 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4241 on TCP/IP sockets, because this apparently causes some broken clients to
4244 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4245 unchanged) from the Cygwin maintainer.
4247 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4248 was supported instead of testing for mbx. Effectively a typo.
4250 15. The spa authenticator server code was not checking that the input it
4251 received was valid base64.
4253 16. The debug output line for the "set" modifier in ACLs was not showing the
4254 name of the variable that was being set.
4256 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4258 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4259 lookup on its own. Something else was needed to trigger the lookup. For
4260 example, a match in host_lookup or the need for a host name in a host list.
4261 Now, if $sender_host_name is referenced and the host name has not yet been
4262 looked up, a lookup is performed. If the lookup fails, the variable remains
4263 empty, and $host_lookup_failed is set to "1".
4265 19. Added "eqi" as a case-independent comparison operator.
4267 20. The saslauthd authentication condition could segfault if neither service
4268 nor realm was specified.
4270 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4271 error message that was logged was misleading, and incoming SMTP
4272 connections were dropped. The message is now more accurate, and temporary
4273 errors are given to SMTP connections.
4275 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4276 (except RSET) with a 421 error, until QUIT is received. However, it was
4277 failing to send a response to QUIT.
4279 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4280 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4281 ACL is now run after the helo_try_verify_hosts code.
4283 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4284 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4285 case-independent, so other case variants are also recognized). Apparently
4286 some systems use these upper case variants.
4288 25. If more than two messages were waiting for the same host, and a transport
4289 filter was specified for the transport, Exim sent two messages over the
4290 same TCP/IP connection, and then failed with "socket operation on non-
4291 socket" when it tried to send the third.
4293 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4295 27. The extern definition of crypt16() in expand.c was not being excluded when
4296 the OS had its own crypt16() function.
4298 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4299 as a preferred synonym for return_size_limit, both as an option and as an
4302 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4304 30. Changed the default configuration ACL to relax the local part checking rule
4305 for addresses that are not in any local domains. For these addresses,
4306 slashes and pipe symbols are allowed within local parts, but the sequence
4307 /../ is explicitly forbidden.
4309 31. SPA server authentication was not clearing the challenge buffer before
4312 32. log_message in a "warn" ACL statement was writing to the reject log as
4313 well as to the main log, which contradicts the documentation and doesn't
4314 seem right (because no rejection is happening). So I have stopped it.
4316 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4317 However, I am unable to do any testing of this.
4319 34. Fixed an infelicity in the appendfile transport. When checking directories
4320 for a mailbox, to see if any needed to be created, it was accidentally
4321 using path names with one or more superfluous leading slashes; tracing
4322 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4324 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4325 contents are added to the log line that is written for every discarded
4326 recipient. (Previously a log_message setting was ignored.)
4328 36. The ${quote: operator now quotes the string if it is empty.
4330 37. The install script runs exim in order to find its version number. If for
4331 some reason other than non-existence or emptiness, which it checks, it
4332 could not run './exim', it was installing it with an empty version number,
4333 i.e. as "exim-". This error state is now caught, and the installation is
4336 38. An argument was missing from the function that creates an error message
4337 when Exim fails to connect to the socket for saslauthd authentication.
4338 This could cause Exim to crash, or give a corrupted message.
4340 39. Added isip, isip4, and isip6 to ${if conditions.
4342 40. The ACL variables $acl_xx are now saved with the message, and can be
4343 accessed later in routers, transports, and filters.
4345 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4346 strings in the file are not string-expanded.
4348 42. If a MAIL command specified a SIZE value that was too large to fit into an
4349 int variable, the check against message_size_limit failed. Such values are
4350 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4351 day this will have to be increased, but I don't think I want to be around
4352 when emails are that large.
4359 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4360 iconv() is not standard in FreeBSD.
4362 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4363 IPv6 enabled. The observed symptom was a segmentation fault on return from
4364 the function os_common_find_running_interfaces() in src/os.c.
4366 3. In the check_special_case() function in daemon.c I had used "errno" as an
4367 argument name, which causes warnings on some systems. This was basically a
4368 typo, since it was named "eno" in the comments!
4370 4. The code that waits for the clock to tick (at a resolution of some fraction
4371 of a second) so as to ensure message-id uniqueness was always waiting for
4372 at least one whole tick, when it could have waited for less. [This is
4373 almost certainly not relevant at current processor speeds, where it is
4374 unlikely to ever wait at all. But we try to future-proof.]
4376 5. The function that sleeps for a time interval that includes fractions of a
4377 second contained a race. It did not block SIGALRM between setting the
4378 timer, and suspending (a couple of lines later). If the interval was short
4379 and the sigsuspend() was delayed until after it had expired, the suspension
4380 never ended. On busy systems this could lead to processes getting stuck for
4383 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4384 process, before it forks any delivery processes. The open lookup caching
4385 mechanism meant that the open file or database connection was passed into
4386 the delivery process. The problem was that delivery processes always tidy
4387 up cached lookup data. This could cause a problem for the next delivery
4388 process started by the queue runner, because the external queue runner
4389 process does not know about the closure. So the next delivery process
4390 still has data in the lookup cache. In the case of a file lookup, there was
4391 no problem because closing a file descriptor in a subprocess doesn't affect
4392 the parent. However, if the lookup was caching a connection to a database,
4393 the connection was closed, and the second delivery process was likely to
4394 see errors such as "PGSQL: query failed: server closed the connection
4395 unexpectedly". The problem has been fixed by closing all cached lookups
4396 in a queue runner before running a delivery process.
4398 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4399 seem to have the "const" qualifier which it has on other OS. I've
4402 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4403 *of the same type* (client or server) with the same public name that an
4404 error should be diagnosed.
4406 9. When Exim looked up a host name for an IP address, but failed to find the
4407 original IP address when looking up the host name (a safety check), it
4408 output the message "<ip address> does not match any IP for NULL", which was
4409 confusing, to say the least. The bug was that the host name should have
4410 appeared instead of "NULL".
4412 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4413 user that is built into the binary, and the -C or -D options is used, root
4414 privilege is dropped before the configuration file is read. In addition,
4415 logging is switched to stderr instead of the normal log files. If the
4416 configuration then re-defines the Exim user, the unprivileged environment
4417 is probably not what is expected, so Exim logs a panic warning message (but
4420 However, if deliver_drop_privilege is set, the unprivileged state may well
4421 be exactly what is intended, so the warning has been cut out in that case,
4422 and Exim is allowed to try to write to its normal log files.
4428 1. smtp_return_error_details was not giving details for temporary sender
4429 or receiver verification errors.
4431 2. Diagnose a configuration error if two authenticators have the same public
4434 3. Exim used not to create the message log file for a message until the first
4435 delivery attempt. This could be confusing when incoming messages were held
4436 for policy or load reasons. The message log file is now created at the time
4437 the message is received, and an initial "Received" line is written to it.
4439 4. The automatically generated man page for command line options had a minor
4440 bug that caused no ill effects; however, a more serious problem was that
4441 the procedure for building the man page automatically didn't always
4442 operate. Consequently, release 4.20 contains an out-of-date version. This
4443 shouldn't happen again.
4445 5. When building Exim with embedded Perl support, the script that builds the
4446 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4447 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4449 6. The freeze_tell option was not being used for messages that were frozen on
4450 arrival, either by an ACL or by local_scan().
4452 7. Added the smtp_incomplete_transaction log selector.
4454 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4455 was accepting AUTH without a new EHLO.
4457 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4458 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4461 10. Logging of TCP/IP connections (when configured) now happens in the main
4462 daemon process instead of the child process, so that the TCP/IP connection
4463 count is more accurate (but it can never be perfect).
4465 11. The use of "drop" in a nested ACL was not being handled correctly in the
4466 outer ACL. Now, if condition failure induced by the nested "drop" causes
4467 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4468 or "require"), the connection is dropped.
4470 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4471 that yield "discard" can now be used with an "accept" or a "discard" verb,
4472 but an error is generated for any others (because I can't see a useful way
4473 to define what should happen).
4475 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4476 are now processed in the same way as lines in the Exim configuration file.
4477 In particular, continuation lines are supported.
4479 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4481 15. Added -ti meaning -t -i.
4483 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4484 domains, and warn by logging if others are found.
4486 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4487 returned from GIFCONF when Exim is trying to find the list of interfaces on
4488 a host. The code in os.c has been modified to copy each ifreq to an aligned
4489 structure in all cases.
4491 Also, in some cases, the returned ifreq's were being copied to a 'struct
4492 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4493 means the last couple of bytes of an IPv6 address could be chopped if the
4494 ifreq contained only a normal sockaddr (14 bytes storage).
4496 18. Named domain lists were not supported in the hosts_treat_as_local option.
4497 An entry such as +xxxx was not recognized, and was treated as a literal
4500 19. Ensure that header lines added by a DATA ACL are included in the reject log
4501 if the ACL subsequently rejects the message.
4503 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4504 MD5 (which is deprecated).
4506 21. When testing a filter file using -bf, Exim was writing a message when it
4507 took the sender from a "From " line in the message, but it was not doing so
4508 when it took $return_path from a Return-Path: header line. It now does.
4510 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4511 with a valid header line field name (a series of printing characters
4512 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4514 23. Changed "disc" in the source to "disk" to conform to the documentation and
4515 the book and for uniformity.
4517 24. Ignore Sendmail's -Ooption=value command line item.
4519 25. When execve() failed while trying to run a command in a pipe transport,
4520 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4521 could be confused with a return value of 69 from the command itself. This
4522 has been changed to 127, the value the shell returns if it is asked to run
4523 a non-existent command. The wording for the related log line suggests a
4524 non-existent command as the problem.
4526 26. If received_header_text expands to an empty string, do not add a Received:
4527 header line to the message. (Well, it adds a token one on the spool, but
4528 marks it "old" so that it doesn't get used or transmitted.)
4530 27. Installed eximstats 1.28 (addition of -nt option).
4532 28. There was no check for failure on the call to getsockname() in the daemon
4533 code. This can fail if there is a shortage of resources on the system, with
4534 ENOMEM, for example. A temporary error is now given on failure.
4536 29. Contrary to the C standard, it seems that in some environments, the
4537 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4538 program. Exim now does this explicitly; it affects the formatting of
4539 timestamps using strftime().
4541 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4542 complaints. I've now initialized all the variables, to avoid this.
4544 32. Header lines added by a system filter were not being "seen" during
4545 transport-time rewrites.
4547 33. The info_callback() function passed to OpenSSL is set up with type void
4548 (*)(SSL *, int, int), as described somewhere. However, when calling the
4549 function (actually a macro) that sets it up, the type void(*)() is
4550 expected. I've put in a cast to prevent warnings from picky compilers.
4552 34. If a DNS black list lookup found a CNAME record, but there were no A
4553 records associated with the domain it pointed at, Exim crashed.
4555 35. If a DNS black list lookup returned more than one A record, Exim ignored
4556 all but the first. It now scans all returned addresses if a particular IP
4557 value is being sought. In this situation, the contents of the
4558 $dnslist_value variable are a list of all the addresses, separated by a
4561 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4562 to accept a host name and all its aliases if the forward lookup for any of
4563 them yielded the IP address of the incoming connection. Now it accepts only
4564 those names whose forward lookup yields the correct IP address. Any other
4565 names are discarded. This closes a loophole whereby a rogue DNS
4566 administrator could create reverse DNS records to break through a
4567 wildcarded host restriction in an ACL.
4569 37. If a user filter or a system filter that ran in a subprocess used any of
4570 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4571 the wrong values were passed to the pipe command ($thisaddress had the
4572 value of $0, $0 had the value of $1, etc). This bug was introduced by
4573 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4575 38. Improved the line breaking for long SMTP error messages from ACLs.
4576 Previously, if there was no break point between 40 and 75 characters, Exim
4577 left the rest of the message alone. Two changes have been made: (a) I've
4578 reduced the minimum length to 35 characters; (b) if it can't find a break
4579 point between 35 and 75 characters, it looks ahead and uses the first one
4580 that it finds. This may give the occasional overlong line, but at least the
4581 remaining text gets split now.
4583 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4584 file descriptors might be low, and that setting 1000 would always raise it.
4585 It turns out that in some environments, the limit is already over 1000 and
4586 that lowering it causes trouble. So now Exim takes care not to decrease it.
4588 40. When delivering a message, the value of $return_path is set to $sender_
4589 address at the start of routing (routers may change the value). By an
4590 oversight, this default was not being set up when an address was tested by
4591 -bt or -bv, which affected the outcome if any router or filter referred to
4594 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4595 log only when "message" or "log_message" are set. However, if one of the
4596 conditions was an address verification, or a call to a nested ACL, the
4597 messages generated by the underlying test were being passed through. This
4598 no longer happens. The underlying message is available in $acl_verify_
4599 message for both "message" and "log_message" expansions, so it can be
4600 passed through if needed.
4602 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4603 new expansion $bh_ to give the encoded byte string without charset
4604 translation. Translation happens only if iconv() is available; HAVE_ICONV
4605 indicates this at build time. HEADERS_CHARSET gives the charset to
4606 translate to; headers_charset can change it in the configuration, and
4607 "headers charset" can change it in an individual filter file.
4609 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4610 that creates RFC 2047 encoded "words" labels them as that charset instead
4611 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4612 expansion operator; (ii) when Exim creates a From: line for a local
4613 message; (iii) when a header line is rewritten to include a "phrase" part.
4615 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4616 buggy, causing it to skip the first lines of messages whose message ID
4617 ended in 'D'. This would not have bitten before Exim release 4.14, because
4618 message IDs were unlikely to end in 'D' before then. The effect was to have
4619 incorrect size information for certain domains.
4621 45. #include "config.h" was missing at the start of the crypt16.c module. This
4622 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4625 46. If there was a timeout during a "random" callout check, Exim treated it as
4626 a failure of the random address, and carried on sending RSET and the real
4627 address. If the delay was just some slowness somewhere, the response to the
4628 original RCPT would be taken as a response to RSET and so on, causing
4629 mayhem of various kinds.
4631 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4632 when I implemented it. It didn't allow for the fact that some option values
4633 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4634 the right test for positive values.
4636 48. Domain names in DNS records are case-independent. Exim always looks them up
4637 in lower case. Some resolvers return domain names in exactly the case they
4638 appear in the zone file, that is, they may contain uppercase letters. Not
4639 all resolvers do this - some return always lower case. Exim was treating a
4640 change of case by a resolver as a change of domain, similar to a widening
4641 of a domain abbreviation. This triggered its re-routing code and so it was
4642 trying to route what was effectively the same domain again. This normally
4643 caused routing to fail (because the router wouldn't handle the domain
4644 twice). Now Exim checks for this case specially, and just changes the
4645 casing of the domain that it ultimately uses when it transmits the message
4648 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4651 50. If a filter generated a file delivery with a non-absolute name (possible if
4652 no home directory exists for the router), the forbid_file option was not
4655 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4656 the existing equality matching.
4658 52. Exim was using ints instead of ino_t variables in some places where it was
4659 dealing with inode numbers.
4661 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4662 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4663 called TMPDIR, and if it finds it is different, it changes its value.
4665 54. The smtp_printf() function is now made available to local_scan() so
4666 additional output lines can be written before returning. There is also an
4667 smtp_fflush() function to enable the detection of a dropped connection.
4668 The variables smtp_input and smtp_batched_input are exported to
4671 55. Changed the default runtime configuration: the message "Unknown user"
4672 has been removed from the ACL, and instead placed on the localuser router,
4673 using the cannot_route_message feature. This means that any verification
4674 failures that generate their own messages won't get overridden. Similarly,
4675 the "Unrouteable address" message that was in the ACL for unverifiable
4676 relay addresses has also been removed.
4678 56. Added hosts_avoid_esmtp to the smtp transport.
4680 57. The exicyclog script was not checking for the esoteric option
4681 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4682 will work only if exicyclog is run under the appropriate euid.
4684 58. Following a discussion on the list, the rules by which Exim recognises line
4685 endings on incoming messages have been changed. The -dropcr and drop_cr
4686 options are now no-ops, retained only for backwards compatibility. The
4687 following line terminators are recognized: LF CRLF CR. However, special
4688 processing applies to CR:
4690 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4691 nor a local message in the state where . is a terminator.
4693 (ii) If a bare CR is encountered in a header line, an extra space is added
4694 after the line terminator so as not to end the header. The reasoning
4695 behind this is that bare CRs in header lines are most likely either
4696 to be mistakes, or people trying to play silly games.
4698 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4699 was being incorrectly given as 18 bytes larger than it should have been.
4700 This is a VOB (very old bug).
4702 60. This may never have affected anything current, but just in case it has:
4703 When the local host is found other than at the start of a list of hosts,
4704 the local host, those with the same MX, and any that follow, are discarded.
4705 When the list in question was part of a longer list of hosts, the following
4706 hosts (not currently being processed) were also being discarded. This no
4707 longer happens. I'm not sure if this situation could ever has previously
4710 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4713 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4714 header. This is something that is recommended in a new Internet Draft, and
4715 is something that is documented as being done by Sendmail. There are two
4716 possible values. For messages generated by the autoreply transport, Exim
4719 Auto-Submitted: auto-replied
4721 whereas for all other generated messages (e.g. bounces) it adds
4723 Auto-Submitted: auto-generated
4725 63. The "personal" condition in filters now includes a test for the
4726 Auto-Submitted: header. If it contains the string "auto-" the message it
4727 not considered personal.
4729 64. Added rcpt_include_affixes as a generic transport option.
4731 65. Added queue_only_override (default true).
4733 66. Added the syslog_duplication option.
4735 67. If what should have been the first header line of a message consisted of
4736 a space followed by a colon, Exim was mis-interpreting it as a header line.
4737 It isn't of course - it is syntactically invalid and should therefore be
4738 treated as the start of the message body. The misbehaviour could have
4739 caused a number of strange effects, including loss of data in subsequent
4740 header lines, and spool format errors.
4742 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4743 client host had authenticated. This control can now be exercised by an ACL
4744 for more flexibility.
4746 69. By default, callouts do not happen when testing with -bh. There is now a
4747 variant, -bhc, which does actually run the callout code, including
4748 consulting and updating the callout cache.
4750 70. Added support for saslauthd authentication, courtesy of Alexander
4753 71. If statvfs() failed on the spool or log directories while checking their
4754 size for availability, Exim confusingly gave the error "space shortage".
4755 Furthermore, in debugging mode it crashed with a floating point exception.
4756 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4757 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4758 serious problem, Exim now writes to the main and panic logs when this
4759 happens, with details of the failure. It then refuses to accept the
4760 incoming message, giving the message "spool directory problem" or "log
4761 directory problem" with a 421 code for SMTP messages.
4763 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4764 0, 1, and 2 exist, because some OS complain for execs without them (see
4765 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4766 descriptors. However, the code omitted to check that the open succeeded,
4767 causing mysterious errors if for some reason the permissions on /dev/null
4768 got screwed. Now Exim writes a message to the main and panic logs, and
4769 bombs out if it can't open /dev/null.
4771 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4772 interact so that it is all more flexible. It is supposed to remain
4773 backwards compatible. Also added extra_local_interfaces.
4775 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4776 to bomb out with an assertion failure - to the client this appears as a
4777 connection drop. This problem occurs in the part of the code that was taken
4778 from the Samba project. Fortunately, the assertion is in a very simple
4779 function, so I have fixed this by reproducing the function inline in the
4780 one place where it is called, and arranging for authentication to fail
4781 instead of killing the process with assert().
4783 75. The SPA client code was not working when the server requested OEM rather
4784 than Unicode encoding.
4786 76. Added code to make require_files with a specific uid setting more usable in
4787 the case where statting the file as root fails - usually a non-root-mounted
4788 NFS file system. When this happens and the failure is EACCES, Exim now
4789 forks a subprocess and does the per-uid checking as the relevant uid.
4791 77. Added process_log_path.
4793 78. If log_file_path was not explicitly set, a setting of check_log_space or
4794 check_log_inodes was ignored.
4796 79. If a space check for the spool or log partitions fails, the incident is now
4797 logged. Of course, in the latter case the data may get lost...
4799 80. Added the %p formatting code to string_format() so that it can be used to
4800 print addresses in debug_print(). Adjusted all the address printing in the
4801 debugging in store.c to use %p rather than %d.
4803 81. There was a concern that a line of code in smtp_in.c could overflow a
4804 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4805 initially expressed, the concern was not well-founded, because trailing
4806 spaces are removed early. However, if the trailing spaces were followed by
4807 a NULL, they did not get removed, so the overflow was possible. Two fixes
4810 (a) I re-wrote the offending code in a cleaner fashion.
4811 (b) If an incoming SMTP command contains a NULL character, it is rejected
4814 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4815 runs initgroups(), so that if the Exim user is in any additional groups,
4816 they will be used during message reception.
4822 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob