1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.532 2007/09/21 08:30:38 steve Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
6 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
7 as mailq or other aliases. Changed the --help handling significantly
8 to do whats expected. exim_usage() emits usage/help information.
10 SC/01 Added the -bylocaldomain option to eximstats.
15 PH/01 Another patch from the Sieve maintainer.
17 PH/02 When an IPv6 address is converted to a string for single-key lookup
18 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
19 dots are used instead of colons so that keys in lsearch files need not
20 contain colons. This was done some time before quoting was made available
21 in lsearch files. However, iplsearch files do require colons in IPv6 keys
22 (notated using the quote facility) so as to distinguish them from IPv4
23 keys. This meant that lookups for IP addresses in host lists did not work
24 for iplsearch lookups.
26 This has been fixed by arranging for IPv6 addresses to be expressed with
27 colons if the lookup type is iplsearch. This is not incompatible, because
28 previously such lookups could never work.
30 The situation is now rather anomolous, since one *can* have colons in
31 ordinary lsearch keys. However, making the change in all cases is
32 incompatible and would probably break a number of configurations.
34 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
37 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
38 conversion specification without a maximum field width, thereby enabling
39 a rogue spamd server to cause a buffer overflow. While nobody in their
40 right mind would setup Exim to query an untrusted spamd server, an
41 attacker that gains access to a server running spamd could potentially
42 exploit this vulnerability to run arbitrary code as the Exim user.
44 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
45 $primary_hostname instead of what libspf2 thinks the hosts name is.
47 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
48 a directory entry by the name of the lookup key. Previously, if a
49 symlink pointed to a non-existing file or a file in a directory that
50 Exim lacked permissions to read, a lookup for a key matching that
51 symlink would fail. Now it is enough that a matching directory entry
52 exists, symlink or not. (Bugzilla 503.)
54 PH/03 The body_linecount and body_zerocount variables are now exported in the
57 PH/04 Added the $dnslist_matched variable.
59 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
60 This means they are set thereafter only if the connection becomes
63 PH/06 Added the client_condition to authenticators so that some can be skipped
64 by clients under certain conditions.
66 PH/07 The error message for a badly-placed control=no_multiline_responses left
67 "_responses" off the end of the name.
69 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
71 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
72 (without spaces) instead of just copying the configuration text.
74 PH/10 Added the /noupdate option to the ratelimit ACL condition.
76 PH/11 Added $max_received_linelength.
78 PH/12 Added +ignore_defer and +include_defer to host lists.
80 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
81 way in which PCRE > 7.0 is built.
83 PH/14 Implemented queue_only_load_latch.
85 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
86 MAIL command. The effect was to mangle the value on 64-bit systems.
88 PH/16 Another patch from the Sieve maintainer.
90 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
92 PH/18 If a system quota error occurred while trying to create the file for
93 a maildir delivery, the message "Mailbox is full" was not appended to the
94 bounce if the delivery eventually timed out. Change 4.67/27 below applied
95 only to a quota excession during the actual writing of the file.
97 PH/19 It seems that peer DN values may contain newlines (and other non-printing
98 characters?) which causes problems in log lines. The DN values are now
99 passed through string_printing() before being added to log lines.
101 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
102 and InterBase are left for another time.)
104 PH/21 Added message_body_newlines option.
106 PH/22 Guard against possible overflow in moan_check_errorcopy().
108 PH/23 POSIX allows open() to be a macro; guard against that.
110 PH/24 If the recipient of an error message contained an @ in the local part
111 (suitably quoted, of course), incorrect values were put in $domain and
112 $local_part during the evaluation of errors_copy.
118 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
119 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
122 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
123 issue a MAIL command.
125 PH/02 In an ACL statement such as
127 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
129 if a client was not listed at all, or was listed with a value other than
130 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
131 the condition was not true (as it should be), so access was not denied.
132 The bug was that the ! inversion was incorrectly passed on to the second
133 item. This has been fixed.
135 PH/03 Added additional dnslists conditions == and =& which are different from
136 = and & when the dns lookup returns more than one IP address.
138 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
139 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
141 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
142 FSYNC, which compiles an option called disable_fsync that allows for
143 bypassing fsync(). The documentation is heavily laced with warnings.
145 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
147 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
148 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
149 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
150 including adding "make clean"; (3) Added -fPIC when compiling the test
151 dynamically loaded module, to get rid of a warning.
153 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
154 message fails, move_frozen_messages = true and ignore_bounce_errors_after
155 = 0s. The bug is otherwise harmless.
157 PH/07 There was a bug in the dovecot authenticator such that the value of
158 $auth1 could be overwritten, and so not correctly preserved, after a
159 successful authentication. This usually meant that the value preserved by
160 the server_setid option was incorrect.
162 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
164 PH/09 Installed PCRE release 7.0.
166 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
167 run for batched SMTP input. It is now run at the start of every message
168 in the batch. While fixing this I discovered that the process information
169 (output by running exiwhat) was not always getting set for -bs and -bS
170 input. This is fixed, and it now also says "batched" for BSMTP.
172 PH/11 Added control=no_pipelining.
174 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
175 patch, slightly modified), and move the expansion of helo_data till after
176 the connection is made in the smtp transport (so it can use these
179 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
181 PH/14 Added log_selector = +pid.
183 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
185 PH/16 Add ${if forany and ${if forall.
187 PH/17 Added dsn_from option to vary the From: line in DSNs.
189 PH/18 Flush SMTP output before performing a callout, unless control =
190 no_callout_flush is set.
192 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
193 was true (the default) a successful delivery failed to delete the retry
194 item, thus causing premature timeout of the address. The bug is now
197 PH/20 Added hosts_avoid_pipelining to the smtp transport.
199 PH/21 Long custom messages for fakedefer and fakereject are now split up
200 into multiline reponses in the same way that messages for "deny" and
201 other ACL rejections are.
203 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
204 with slight modification.
206 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
207 draft, changing the syntax and factoring some duplicate code".
209 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
210 for deliveries of the second and subsequent messages over the same SMTP
213 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
214 ${reduce, with only minor "tidies".
216 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
218 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
219 expansion side effects.
221 PH/27 When a message times out after an over-quota error from an Exim-imposed
222 quota, the bounce message says "mailbox is full". This message was not
223 being given when it was a system quota that was exceeded. It now should
226 MH/03 Made $recipients available in local_scan(). local_scan() already has
227 better access to the recipient list through recipients_list[], but
228 $recipients can be useful in postmaster-provided expansion strings.
230 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
231 in the case of a MAIL command with additional options following the
232 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
233 were accidentally chopped off.
235 PH/29 SMTP synchronization checks are implemented when a command is read -
236 there is a check that no more input is waiting when there shouldn't be
237 any. However, for some commands, a delay in an ACL can mean that it is
238 some time before the response is written. In this time, more input might
239 arrive, invalidly. So now there are extra checks after an ACL has run for
240 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
241 pipelining has not been advertised.
243 PH/30 MH's patch to allow iscntrl() characters to be list separators.
245 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
246 returned in the SMTP response when smtp_return_error_details was false.
249 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
250 instead of the C I/O that was originally supplied, because problems were
253 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
254 Exim which did not show up earlier: it was assuming that a call to
255 SSL_CTX_set_info_callback() might give an error value. In fact, there is
256 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
257 was a macro that became an assignment, so it seemed to work. This has
258 changed to a proper function call with a void return, hence the compile
259 error. Exim's code has been fixed.
261 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
264 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
266 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
268 PH/37 If a message is not accepted after it has had an id assigned (e.g.
269 because it turns out to be too big or there is a timeout) there is no
270 "Completed" line in the log. When some messages of this type were
271 selected by exigrep, they were listed as "not completed". Others were
272 picked up by some special patterns. I have improved the selection
273 criteria to be more general.
275 PH/38 The host_find_failed option in the manualroute router can now be set
276 to "ignore", to completely ignore a host whose IP address cannot be
277 found. If all hosts are ignored, the behaviour is controlled by the new
278 host_all_ignored option.
280 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
281 homing or because of multiple MX records with /mx) generated more than
282 one IP address, and the following item turned out to be the local host,
283 all the secondary addresses of the first item were incorrectly removed
284 from the list, along with the local host and any following hosts (which
285 is what is supposed to happen).
287 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
288 whoever called Exim into the -H file. In the case of the daemon it was
289 behaving confusingly. When first started, it used values for whoever
290 started the daemon, but after a SIGHUP it used the Exim user (because it
291 calls itself on a restart). I have changed the code so that it now always
294 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
295 message are rejected with the same error (e.g. no authentication or bad
296 sender address), and a DATA command is nevertheless sent (as can happen
297 with PIPELINING or a stupid MUA), the error message that was given to the
298 RCPT commands is included in the rejection of the DATA command. This is
299 intended to be helpful for MUAs that show only the final error to their
302 PH/42 Another patch from the Sieve maintainer.
304 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
305 Eximstats - Fixed some broken HTML links and added missing column headers
307 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
308 columns for Rejects, Temp Rejects, Ham, and Spam rows.
310 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
312 PH/43 Yet another patch from the Sieve maintainer.
314 PH/44 I found a way to check for a TCP/IP connection going away before sending
315 the response to the final '.' that terminates a message, but only in the
316 case where the client has not sent further data following the '.'
317 (unfortunately, this is allowed). However, in many cases there won't be
318 any further data because there won't be any more messages to send. A call
319 to select() can be used: if it shows that the input is "ready", there is
320 either input waiting, or the socket has been closed. An attempt to read
321 the next input character can distinguish the two cases. Previously, Exim
322 would have sent an OK response which the client would never have see.
323 This could lead to message repetition. This fix should cure that, at
324 least in a lot of common cases.
326 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
327 advertised in response to EHLO.
333 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
334 fixed by 4.65/MH/01 (is this a record?) are fixed:
336 (i) An empty string was always treated as zero by the numeric comparison
337 operators. This behaviour has been restored.
339 (ii) It is documented that the numeric comparison operators always treat
340 their arguments as decimal numbers. This was broken in that numbers
341 starting with 0 were being interpreted as octal.
343 While fixing these problems I realized that there was another issue that
344 hadn't been noticed. Values of message_size_limit (both the global option
345 and the transport option) were treated as octal if they started with 0.
346 The documentation was vague. These values are now always treated as
347 decimal, and I will make that clear in the documentation.
353 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
354 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
357 MH/01 Don't check that the operands of numeric comparison operators are
358 integers when their expansion is in "skipping" mode (fixes bug
359 introduced by 4.64-PH/07).
361 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
362 child addresses, Exim now panics and dies. Previously, because the count
363 is held in a short int, deliveries were likely to be lost. As such a
364 large number of recipients for a single message is ridiculous
365 (performance will be very, very poor), I have chosen to impose a limit
366 rather than extend the field.
372 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
373 leftover -K file (the existence of which was triggered by #402).
374 While we were at it, introduced process PID as part of the -K
375 filename. This should rule out race conditions when creating
378 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
379 processing considerably. Previous code took too long for large mails,
380 triggering a timeout which in turn triggers #401.
382 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
383 in the DK code in transports.c. sendfile() is not really portable,
384 hence the _LINUX specificness.
386 TF/01 In the add_headers option to the mail command in an Exim filter,
387 there was a bug that Exim would claim a syntax error in any
388 header after the first one which had an odd number of characters
391 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
392 callout verification, Exim cached a "reject" for the entire domain. This
393 is correct for most verifications, but it is not correct for a recipient
394 verification with use_sender or use_postmaster set, because in that case
395 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
396 case of MAIL FROM:<> rejection from other early rejections (e.g.
397 rejection of HELO). When verifying a recipient using a non-null MAIL
398 address, the cache is ignored if it shows MAIL FROM:<> rejection.
399 Whatever the result of the callout, the value of the domain cache is
400 left unchanged (for any other kind of callout, getting as far as trying
401 RCPT means that the domain itself is ok).
403 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
406 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
407 manifest itself as EPIPE rather than ECONNECT. When tidying away a
408 session, the daemon ignores ECONNECT errors and logs others; it now
409 ignores EPIPE as well.
411 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
412 (quoted-printable decoding).
414 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
415 later the small subsequent patch to fix an introduced bug.
417 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
419 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
421 PH/08 An error is now given if message_size_limit is specified negative.
423 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
424 to be given (somewhat) arbitrary names.
426 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
429 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
430 miscellaneous code fixes
432 PH/10 Added the log_reject_target ACL modifier to specify where to log
435 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
436 hostname. This is wrong, because it relates to the incoming message (and
437 probably the interface on which it is arriving) and not to the outgoing
438 callout (which could be using a different interface). This has been
439 changed to use the value of the helo_data option from the smtp transport
440 instead - this is what is used when a message is actually being sent. If
441 there is no remote transport (possible with a router that sets up host
442 addresses), $smtp_active_hostname is used.
444 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
445 tweaks were necessary in order to get it to work (see also 21 below):
446 (a) The code assumed that strncpy() returns a negative number on buffer
447 overflow, which isn't the case. Replaced with Exim's string_format()
449 (b) There were several signed/unsigned issues. I just did the minimum
450 hacking in of casts. There is scope for a larger refactoring.
451 (c) The code used strcasecmp() which is not a standard C function.
452 Replaced with Exim's strcmpic() function.
453 (d) The code set only $1; it now sets $auth1 as well.
454 (e) A simple test gave the error "authentication client didn't specify
455 service in request". It would seem that Dovecot has changed its
456 interface. Fortunately there's a specification; I followed it and
457 changed what the client sends and it appears to be working now.
459 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
462 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
463 address A is aliased to B and C, where B exists and C does not. Without
464 -v the output is "A verified" because verification stops after a
465 successful redirection if more than one address is generated. However,
466 with -v the child addresses are also verified. Exim was outputting "A
467 failed to verify" and then showing the successful verification for C,
468 with its parentage. It now outputs "B failed to verify", showing B's
469 parentage before showing the successful verification of C.
471 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
472 look up a TXT record in a specific list after matching in a combined
475 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
476 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
477 they consult the DNS. I had assumed they would set it the way they
478 wanted; and indeed my experiments on Linux seem to show that in some
479 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
480 To be on the safe side, however, I have now made the interface to
481 host_find_byname() similar to host_find_bydns(), with an argument
482 containing the DNS resolver options. The host_find_byname() function now
483 sets these options at its start, just as host_find_bydns() does. The smtp
484 transport options dns_qualify_single and dns_search_parents are passed to
485 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
486 of host_find_byname() use the default settings of RES_DEFNAMES
487 (qualify_single) but not RES_DNSRCH (search_parents).
489 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
490 spool_read_header() do less string testing, by means of a preliminary
491 switch on the second character of optional "-foo" lines. (This is
492 overdue, caused by the large number of possibilities that now exist.
493 Originally there were few.) While I was there, I also converted the
494 str(n)cmp tests so they don't re-test the leading "-" and the first
495 character, in the hope this might squeeze out yet more improvement.
497 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
498 flag allowing group syntax was set by the header_syntax check but not
499 turned off, possible causing trouble later; (2) The flag was not being
500 set at all for the header_verify test, causing "group"-style headers to
501 be rejected. I have now set it in this case, and also caused header_
502 verify to ignore an empty address taken from a group. While doing this, I
503 came across some other cases where the code for allowing group syntax
504 while scanning a header line wasn't quite right (mostly, not resetting
505 the flag correctly in the right place). These bugs could have caused
506 trouble for malformed header lines. I hope it is now all correct.
508 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
509 with the "reply" argument non-NULL. The code, however (which originally
510 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
511 but it didn't always do it. This confused somebody who was copying the
512 code for some other use. I have removed all the tests.
514 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
515 feature that was used to support insecure browsers during the U.S. crypto
516 embargo. It requires special client support, and Exim is probably the
517 only MTA that supported it -- and would never use it because real RSA is
518 always available. This code has been removed, because it had the bad
519 effect of slowing Exim down by computing (never used) parameters for the
520 RSA_EXPORT functionality.
522 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
523 authenticator to fail if there's a tab character in the incoming data
524 (there should never be unless someone is messing about, as it's supposed
525 to be base64-encoded). Also added, on Timo's advice, the "secured" option
526 if the connection is using TLS or if the remote IP is the same as the
527 local IP, and the "valid-client-cert option" if a client certificate has
530 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
531 authenticators. This can be used for authorization after authentication
532 succeeds. (In the case of plaintext, it servers for both authentication
535 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
536 if any retry times were supplied.
538 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
539 connection, where there is no client IP address to check. In this
540 situation, the verify now always succeeds.
542 PH/25 Applied John Jetmore's -Mset patch.
544 PH/26 Added -bem to be like -Mset, but loading a message from a file.
546 PH/27 In a string expansion for a processed (not raw) header when multiple
547 headers of the same name were present, leading whitespace was being
548 removed from all of them, but trailing whitespace was being removed only
549 from the last one. Now trailing whitespace is removed from each header
550 before concatenation. Completely empty headers in a concatenation (as
553 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
554 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
556 PH/29 [Removed. This was a change that I later backed out, and forgot to
557 correct the ChangeLog entry (that I had efficiently created) before
558 committing the later change.]
560 PH/30 Exim was sometimes attempting to deliver messages that had suffered
561 address errors (4xx response to RCPT) over the same connection as other
562 messages routed to the same hosts. Such deliveries are always "forced",
563 so retry times are not inspected. This resulted in far too many retries
564 for the affected addresses. The effect occurred only when there were more
565 hosts than the hosts_max_try setting in the smtp transport when it had
566 the 4xx errors. Those hosts that it had tried were not added to the list
567 of hosts for which the message was waiting, so if all were tried, there
568 was no problem. Two fixes have been applied:
570 (i) If there are any address or message errors in an SMTP delivery, none
571 of the hosts (tried or untried) are now added to the list of hosts
572 for which the message is waiting, so the message should not be a
573 candidate for sending over the same connection that was used for a
574 successful delivery of some other message. This seems entirely
575 reasonable: after all the message is NOT "waiting for some host".
576 This is so "obvious" that I'm not sure why it wasn't done
577 previously. Hope I haven't missed anything, but it can't do any
578 harm, as the worst effect is to miss an optimization.
580 (ii) If, despite (i), such a delivery is accidentally attempted, the
581 routing retry time is respected, so at least it doesn't keep
582 hammering the server.
584 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
585 in ${readsocket because some servers need this prod.
587 PH/32 Added some extra debug output when updating a wait-xxx database.
589 PH/33 The hint "could be header name not terminated by colon", which has been
590 given for certain expansion errors for a long time, was not being given
591 for the ${if def:h_colon_omitted{... case.
593 PH/34 The spec says: "With one important exception, whenever a domain list is
594 being scanned, $domain contains the subject domain." There was at least
595 one case where this was not true.
597 PH/35 The error "getsockname() failed: connection reset by peer" was being
598 written to the panic log as well as the main log, but it isn't really
599 panic-worthy as it just means the connection died rather early on. I have
600 removed the panic log writing for the ECONNRESET error when getsockname()
603 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
604 runs only) independently of the message's sender address. This meant
605 that, if the 4xx error was in fact related to the sender, a different
606 message to the same recipient with a different sender could confuse
607 things. In particualar, this can happen when sending to a greylisting
608 server, but other circumstances could also provoke similar problems.
609 I have changed the default so that the retry time for these errors is now
610 based a combination of the sender and recipient addresses. This change
611 can be overridden by setting address_retry_include_sender=false in the
614 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
615 remote server are returned as part of bounce messages. This was not
616 happening for LMTP over a pipe (the lmtp transport), but now it is the
617 same for both kinds of LMTP.
619 PH/38 Despite being documented as not happening, Exim was rewriting addresses
620 in header lines that were in fact CNAMEs. This is no longer the case.
622 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
623 and queue runs started by the daemon processed all messages. This has
624 been fixed so that -R and -S can now usefully be given with -q<time>.
626 PH/40 Import PCRE release 6.7 (fixes some bugs).
628 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
630 PH/42 Give an error if -q is specified more than once.
632 PH/43 Renamed the variables $interface_address and $interface_port as
633 $received_ip_address and $received_port, to make it clear that these
634 values apply to message reception, and not to the outgoing interface when
635 a message is delivered. (The old names remain recognized, of course.)
637 PH/44 There was no timeout on the connect() call when using a Unix domain
638 socket in the ${readsocket expansion. There now is.
640 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
641 be meaningful with "accept".
643 SC/01 Eximstats V1.43
644 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
646 SC/02 Eximstats V1.44
647 Use a glob alias rather than an array ref in the generated
648 parser. This improves both readability and performance.
650 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
651 Collect SpamAssassin and rejection statistics.
652 Don't display local sender or destination tables unless
653 there is data to show.
654 Added average volumes into the top table text output.
656 SC/04 Eximstats V1.46
657 Collect data on the number of addresses (recipients)
658 as well as the number of messages.
660 SC/05 Eximstats V1.47
661 Added 'Message too big' to the list of mail rejection
662 reasons (thanks to Marco Gaiarin).
664 SC/06 Eximstats V1.48
665 Mainlog lines which have GMT offsets and are too short to
666 have a flag are now skipped.
668 SC/07 Eximstats V1.49 (Alain Williams)
669 Added the -emptyok flag.
671 SC/08 Eximstats V1.50
672 Fixes for obtaining the IP address from reject messages.
674 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
675 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
676 whitesspace changes from 4.64-PH/27
678 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
681 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
682 are found, allow negative numbers in numeric criteria)
684 JJ/06 exipick.20061117.2, added new $message_body_missing variable
686 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
687 to match changes made in 4.64-PH/43
689 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
691 PH/47 Put in an explicit test for a DNS lookup of an address record where the
692 "domain" is actually an IP address, and force a failure. This locks out
693 those revolvers/nameservers that support "A-for-A" lookups, in
694 contravention of the specifications.
696 PH/48 When a host name was looked up from an IP address, and the subsequent
697 forward lookup of the name timed out, the host name was left in
698 $sender_host_name, contrary to the specification.
700 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
701 restricted to single-key lookups, Exim was not diagnosing an error if
702 * or *@ was used with a query-style lookup.
704 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
706 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
707 long ago, but noone interested enough thought of it. Let's just say that
708 the "1.1" means that there are some new functions that weren't there at
709 some point in the past.
711 PH/51 Error processing for expansion failure of helo_data from an smtp
712 transport during callout processing was broken.
714 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
715 tested/used via the -bh/-bhc/-bs options.
717 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
718 bug, fixed in subsequent PCRE releases).
720 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
721 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
723 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
729 SC/01 Use a glob alias rather than an array ref in eximstats generated
730 parser. This improves both readability and performance.
732 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
733 Don't display local sender or destination tables in eximstats unless
734 there is data to show.
735 Added average volumes into the eximstats top table text output.
737 SC/03 Collect data on the number of addresses (recipients) as well
738 as the number of messages in eximstats.
740 TF/01 Correct an error in the documentation for the redirect router. Exim
741 does (usually) call initgroups() when daemonizing.
743 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
744 with consistent privilege compared to when running as a daemon.
746 TF/03 Note in the spec that $authenticated_id is not set for local
747 submissions from trusted users.
749 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
750 Thanks to Dean Brooks <dean@iglou.com> for the patch.
752 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
753 by adding some example configuration directives to the default
754 configuration file. A little bit of work is required to uncomment the
755 directives and define how usernames and passwords are checked, but
756 there is now a framework to start from.
758 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
759 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
760 without this. I don't know how relevant this is to other LDAP libraries.
762 PH/02 Add the verb name to the "unknown ACL verb" error.
764 PH/03 Magnus Holmgren's patch for filter_prepend_home.
766 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
768 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
769 directory not expanded when it should be if an expanded home directory
770 was set for the address (which is overridden by the transport).
772 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
775 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
776 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
777 because it is too late at that time, and has no effect.
779 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
780 security issue with \' (bugzilla #107). I could not use the
781 PQescapeStringConn() function, because it needs a PGconn value as one of
784 PH/08 When testing addresses using -bt, indicate those final addresses that
785 are duplicates that would not cause an additional delivery. At least one
786 person was confused, thinking that -bt output corresponded to deliveries.
787 (Suppressing duplicates isn't a good idea as you lose the information
788 about possibly different redirections that led to the duplicates.)
790 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
791 systems where poll() doesn't work, in particular OS X.
793 PH/10 Added more information to debugging output for retry time not reached.
795 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
796 operations in malware.c.
798 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
801 PH/13 If write_rejectlog was set false when logging was sent to syslog with
802 syslog_duplication set false, log lines that would normally be written
803 both the the main log and to the reject log were not written to syslog at
806 PH/14 In the default configuration, change the use of "message" in ACL warn
807 statements to "add_header".
809 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
810 not followed by a command (e.g. "seen endif").
812 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
813 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
816 PH/17 Added extra conditions to the default value of delay_warning_condition
820 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
821 { match{$h_precedence:}{(?i)bulk|list|junk} } \
822 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
825 The Auto-Submitted: and various List- headers are standardised, whereas I
826 don't think Precedence: ever was.
828 PH/18 Refactored debugging code in route_finduser() to show more information,
829 in particular, the error code if getpwnam() issues one.
831 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
832 This is apparently needed in addition to the PH/07 change above to avoid
833 any possible encoding problems.
835 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
836 but not after initializing Perl.
838 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
839 output them only if debugging. By default they are written stderr,
840 apparently, which is not desirable.
842 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
845 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
848 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
850 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
851 authenticated or an ident call has been made. Suppress the default
852 values for $authenticated_id and $authenticated_sender (but permit -oMai
853 and -oMas) when testing with -bh.
855 PH/24 Re-jigged the order of the tests in the default configuration so that the
856 tests for valid domains and recipients precede the DNS black list and CSA
857 tests, on the grounds that those ones are more expensive.
859 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
860 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
861 This bug exists in every version of Exim that I still have, right back to
864 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
865 However, an attempt to turn on case-sensitivity in a regex key by
866 including (?-i) didn't work because the subject string was already
867 lowercased, and the effects were non-intuitive. It turns out that a
868 one-line patch can be used to allow (?-i) to work as expected.
874 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
875 other effects) broke the use of negated acl sub-conditions.
877 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
880 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
881 "Deny" causes Exim to reject the incoming connection with a 554 error.
882 Unfortunately, if there is a major crisis, such as a disk failure,
883 tcp-wrappers gives "deny", whereas what one would like would be some
884 kind of temporary error. A kludge has been added to help with this.
885 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
886 554 error is used if errno is still zero or contains ENOENT (which occurs
887 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
890 PH/03 Add -lutil to the default FreeBSD LIBS setting.
892 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
893 errors. Otherwise a message that provokes a temporary error (when other
894 messages do not) can cause a whole host to time out.
896 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
897 addresses were routed directly to files or pipes from a redirect router.
898 File deliveries just didn't batch; pipe deliveries might have suffered
901 PH/06 A failure to get a lock for a hints database would erroneously always say
902 "Failed to get write lock", even when it was really a read lock.
904 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
905 of 0600. This has been changed to use the value of the lockfile_mode
906 option (which defaults to 0600).
908 PH/08 Applied small patch from the Sieve maintainer.
910 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
911 folder from quota calculations, a direct delivery into this folder messed
912 up the contents of the maildirsize file. This was because the regex was
913 used only to exclude .Trash (or whatever) when the size of the mailbox
914 was calculated. There was no check that a delivery was happening into an
915 excluded directory. This bug has been fixed by ignoring all quota
916 processing for deliveries into excluded directories.
918 PH/10 Added the maildirfolder_create_regex option to appendfile.
924 PH/01 The code for finding all the local interface addresses on a FreeBSD
925 system running IPv6 was broken. This may well have applied to all BSD
926 systems, as well as to others that have similar system calls. The broken
927 code found IPv4 interfaces correctly, but gave incorrect values for the
928 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
929 that it would not match correctly against @[] and not recognize the IPv6
932 PH/02 The ipliteral router was not recognizing addresses of the form user@
933 [ipv6:....] because it didn't know about the "ipv6:" prefix.
935 PH/03 Added disable_ipv6.
937 PH/04 Changed $reply_address to use the raw form of the headers instead of the
938 decoded form, because it is most often used to construct To: headers
939 lines in autoreplies, and the decoded form may well be syntactically
940 invalid. However, $reply_address has leading white space removed, and all
941 newlines turned into spaces so that the autoreply transport does not
944 PH/05 If group was specified without a user on a router, and no group or user
945 was specified on a transport, the group from the router was ignored.
947 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
948 for visible compile-time settings that can be used to change these
949 numbers, for those that want even more. Backwards compatibility with old
950 spool files has been maintained. However, going back to a previous Exim
951 release will lost any variables that are in spool files.
953 PH/07 Two small changes when running in the test harness: increase delay when
954 passing a TCP/IP connection to a new process, in case the original
955 process has to generate a bounce, and remove special handling of
956 127.0.0.2 (sic), which is no longer necessary.
958 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
959 be the same on different OS.
961 PH/09 Moved a debug statement in filter processing to avoid a race problem when
964 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
965 whether --show-vars was specified or not
967 JJ/02 exipick: Added support for new ACL variable spool format introduced
970 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
971 syntactically invalid From: or Reply-to: line, and a filter used this to
972 generate an autoreply, and therefore failed to obtain an address for the
973 autoreply, Exim could try to deliver to a non-existent relative file
974 name, causing unrelated and misleading errors. What now happens is that
975 it logs this as a hard delivery error, but does not attempt to create a
978 PH/11 The exinext utility has a -C option for testing purposes, but although
979 the given file was scanned by exinext itself; it wasn't being passed on
980 when Exim was called.
982 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
983 an end-of-file indication when reading a command response.
985 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
986 compiled. In many other places in Exim, IPv6 addresses are always
987 recognized, so I have changed this. It also means that IPv4 domain
988 literals of the form [IPV4:n.n.n.n] are now always recognized.
990 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
991 used if the router is not running as root, for example, when verifying at
992 ACL time, or when using -bh. The debugging output from this situation was
993 non-existent - all you got was a failure to exec. I have made two
996 (a) Failures to set uid/gid, the current directory, or a process leader
997 in a subprocess such as that created by queryprogram now generate
998 suitable debugging ouput when -d is set.
1000 (b) The queryprogram router detects when it is not running as root,
1001 outputs suitable debugging information if -d is set, and then runs
1002 the subprocess without attempting to change uid/gid.
1004 PH/15 Minor change to Makefile for building test_host (undocumented testing
1007 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1008 additional section of a DNS packet that returns MX or SRV records.
1009 Instead, it always explicitly searches for A/AAAA records. This avoids
1010 major problems that occur when a DNS server includes only records of one
1011 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1012 fixed another bug: if SRV records were looked up and the corresponding
1013 address records were *not* found in the additional section, the port
1014 values from the SRV records were lost.
1016 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1017 using the correct key (the original address) when searching the retry
1018 rules in order to find which one to use for generating the retry hint.
1020 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1021 adding the default one. Similarly, if it contains a Reply-To: header, the
1022 errors_reply_to option, if set, is not used.
1024 PH/19 When calculating a retry time, Exim used to measure the "time since
1025 failure" by looking at the "first failed" field in the retry record. Now
1026 it does not use this if it is later than than the arrival time of the
1027 message. Instead it uses the arrival time. This makes for better
1028 behaviour in cases where some deliveries succeed, thus re-setting the
1029 "first failed" field. An example is a quota failure for a huge message
1030 when small messages continue to be delivered. Without this change, the
1031 "time since failure" will always be short, possible causing more frequent
1032 delivery attempts for the huge message than are intended.
1033 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1035 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1036 $1, $2, $3) because the numerical variables can be reset during some
1037 expansion items (e.g. "match"), thereby losing the authentication data.
1039 PH/21 Make -bV show the size of off_t variables so that the test suite can
1040 decide whether to run tests for quotas > 2G.
1042 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1043 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1044 filecount value is greater than 2G or if a quota value is greater than 2G
1045 on a system where the size of off_t is not greater than 4, a panic error
1048 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1049 never match. The debug and -bh output now contains an explicit error
1050 message indicating a malformed IPv4 address or mask.
1052 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1053 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1054 PH/23 above applies.
1056 PH/25 Do not write to syslog when running in the test harness. The only
1057 occasion when this arises is a failure to open the main or panic logs
1058 (for which there is an explicit test).
1060 PH/26 Added the /no_tell option to "control=freeze".
1062 PH/27 If a host name lookup failed very early in a connection, for example, if
1063 the IP address matched host_lookup and the reverse lookup yielded a name
1064 that did not have a forward lookup, an error message of the form "no IP
1065 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1066 could be logged. Now it outputs the IP address instead of "NULL".
1068 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1069 allows the sender and the authenticated sender to be set when
1070 submitting a message from within Exim. Since child_open_exim() is
1071 documented for local_scan(), the new function should be too.
1073 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1074 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1075 results in an empty string is now treated as unset.
1077 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1079 PH/31 Added sender_verify_fail logging option.
1081 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1082 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1083 tidied the source and removed it altogether.
1085 PH/33 When a queue run was abandoned because the load average was too high, a
1086 log line was always written; now it is written only if the queue_run log
1087 selector is set. In addition, the log line for abandonment now contains
1088 information about the queue run such as the pid. This is always present
1089 in "start" and "stop" lines but was omitted from the "abandon" line.
1091 PH/34 Omit spaces between a header name and the colon in the error message that
1092 is given when verify = headers_syntax fails (if there are lots of them,
1093 the message gets confusing).
1095 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1096 names, as there are now some PTR records that contain slashes. This check
1097 is only to protect against broken name servers that fall over on strange
1098 characters, so the fact that it applies to all lookups doesn't matter.
1100 PH/36 Now that the new test suite is complete, we can remove some of the
1101 special code in Exim that was needed for the old test suite. For example,
1102 sorting DNS records because real resolvers return them in an arbitrary
1103 order. The new test suite's fake resolver always returns records in the
1106 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1107 bounces) except when queue_only is set, to avoid logging races between
1108 the different processes.
1110 PH/38 Panic-die if .include specifies a non-absolute path.
1112 PH/39 A tweak to the "H" retry rule from its user.
1114 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1115 a label. They prevented compilation on older perls.
1117 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1118 a warning to be raised on newish perls.
1120 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1121 on queue. Changes to match documented behaviour of showing count of
1122 messages matching specified criteria.
1124 PH/40 Changed the default ident timeout from 30s to 5s.
1126 PH/41 Added support for the use of login_cap features, on those BSD systems
1127 that have them, for controlling the resources used by pipe deliveries.
1129 PH/42 The content-scanning code uses fopen() to create files in which to put
1130 message data. Previously it was not paying any attention to the mode of
1131 the files. Exim runs with umask(0) because the rest of the code creates
1132 files with open(), and sets the required mode explicitly. Thus, these
1133 files were ending up world-writeable. This was not a big issue, because,
1134 being within the spool directory, they were not world-accessible. I have
1135 created a function called modefopen, which takes an additional mode
1136 argument. It sets umask(777), creates the file, chmods it to the required
1137 mode, then resets the umask. All the relevant calls to fopen() in the
1138 content scanning code have been changed to use this function.
1140 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1141 to 24 hours. This avoids potential overflow problems when processing G
1142 and H retry rules. I suspect nobody ever tinkers with this value.
1144 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1146 PH/45 When the plaintext authenticator is running as a client, the server's
1147 challenges are checked to ensure they are valid base64 strings. By
1148 default, the authentication attempt is cancelled if an invalid string is
1149 received. Setting client_ignore_invalid_base64 true ignores these errors.
1150 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1151 they are received. Thus, the responses can be made to depend on the
1152 challenges. If an invalid string is ignored, an empty string is placed in
1155 PH/46 Messages that are created by the autoreply transport now contains a
1156 References: header, in accordance with RFCs 2822 and 3834.
1158 PH/47 Added authenticated_sender_force to the smtp transport.
1160 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1162 PH/49 Installed latest patch from the Sieve maintainer.
1164 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1165 was also set, the appendfile transport was unnecessarily scanning a
1166 directory of message files (e.g. for maildir delivery) to find the count
1167 of files (along with the size), even though it did not need this
1168 information. It now does the scan only if it needs to find either the
1169 size of the count of files.
1171 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1173 PH/52 Two bugs concerned with error handling when the smtp transport is
1176 (i) Exim was not creating retry information for temporary errors given
1177 for individual recipients after the DATA command when the smtp transport
1178 was used in LMTP mode. This meant that they could be retried too
1179 frequently, and not timed out correctly.
1181 (ii) Exim was setting the flag that allows error details to be returned
1182 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1183 recipients that were returned after the DATA command.
1185 PH/53 This is related to PH/52, but is more general: for any failing address,
1186 when detailed error information was permitted to be returned to the
1187 sender, but the error was temporary, then after the final timeout, only
1188 "retry timeout exceeded" was returned. Now it returns the full error as
1189 well as "retry timeout exceeded".
1191 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1192 do this, and (what is worse) MTAs that accept it.
1194 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1195 will now be deprecated.
1197 PH/56 New os.c-cygwin from the Cygwin maintainer.
1199 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1200 formats (previously only available in exim formats via -bpr, -bpru,
1201 and -bpra. Now also available in native and exiqgrep formats)
1203 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1204 with very large, slow to parse queues
1206 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1208 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1210 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1211 responses to authentication challenges, though it was showing the
1212 challenges; (ii) I've removed the CR characters from the debug output for
1215 PH/58 Allow for the insertion of a newline as well as a space when a string
1216 is turned into more than one encoded-word during RFC 2047 encoding. The
1217 Sieve code now uses this.
1219 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1220 data_4xx, lost_connection, tls_required.
1222 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1223 message was being sent as an SMTP response.
1225 PH/61 Add -l and -k options to exicyclog.
1227 PH/62 When verifying, if an address was redirected to one new address, so that
1228 verification continued, and the new address failed or deferred after
1229 having set something in $address_data, the value of $address_data was not
1230 passed back to the ACL. This was different to the case when no
1231 redirection occurred. The value is now passed back in both cases.
1233 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1234 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1235 use login_cap.h, so on its own it isn't the distinguishing feature. The
1236 new name refers directly to the setclassresources() function.
1238 PH/65 Added configuration files for NetBSD3.
1240 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1242 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1243 is preferred over IPv4.
1245 PH/68 The bounce_return_message and bounce_return_body options were not being
1246 honoured for bounces generated during the reception of non-SMTP messages.
1247 In particular, this applied to messages rejected by the ACL. This bug has
1248 been fixed. However, if bounce_return_message is true and bounce_return_
1249 body is false, the headers that are returned for a non-SMTP message
1250 include only those that have been read before the error was detected.
1251 (In the case of an ACL rejection, they have all been read.)
1253 PH/69 The HTML version of the specification is now built in a directory called
1254 spec_html instead of spec.html, because the latter looks like a path with
1255 a MIME-type, and this confuses some software.
1257 PH/70 Catch two compiler warnings in sieve.c.
1259 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1260 function verify_get_ident() calls ip_connect() to connect a socket, but
1261 if the "connect()" function timed out, ip_connect() used to close the
1262 socket. However, verify_get_ident() also closes the socket later, and in
1263 between Exim writes to the log, which may get opened at this point. When
1264 the socket was closed in ip_connect(), the log could get the same file
1265 descriptor number as the socket. This naturally causes chaos. The fix is
1266 not to close the socket in ip_connect(); the socket should be closed by
1267 the function that creates it. There was only one place in the code where
1268 this was missing, in the iplookup router, which I don't think anybody now
1269 uses, but I've fixed it anyway.
1271 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1272 well as to direct DNS lookups. Otherwise the handling of names in host
1273 lists is inconsistent and therefore confusing.
1279 PH/01 Two changes to the default runtime configuration:
1281 (1) Move the checks for relay_from_hosts and authenticated clients from
1282 after to before the (commented out) DNS black list checks.
1284 (2) Add control=submission to the relay_from_hosts and authenticated
1285 clients checks, on the grounds that messages accepted by these
1286 statements are most likely to be submissions.
1288 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1290 (1) Generate an error if the third argument for the ${prvs expansion is
1293 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1296 (3) Reset the variables that are obtained from the first argument of
1297 ${prvscheck and used in the second argument before leaving the code,
1298 because their memory is reclaimed, so using them afterwards may do
1301 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1302 one (it's much easier than Tom thought :-).
1304 (5) Because of (4), we can now allow for the use of $prvscheck_result
1305 inside the third argument.
1307 PH/03 For some reason, the default setting of PATH when running a command from
1308 a pipe transport was just "/usr/bin". I have changed it to
1311 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1312 anything to be listed in the output from -bV.
1314 PH/05 When a filter generated an autoreply, the entire To: header line was
1315 quoted in the delivery log line, like this:
1317 => >A.N.Other <ano@some.domain> <original@ddress> ...
1319 This has been changed so that it extracts the operative address. There
1320 may be more than one such address. If so, they are comma-separated, like
1323 => >ano@some.domain,ona@other.domain <original@ddress> ...
1325 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1326 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1327 not being looked up in the rDNS to get a host name, Exim was showing the
1328 IP address twice in Received: lines, even though the IP addresses were
1329 identical. For example:
1331 Received: from [1.2.3.4] (helo=[1.2.3.4])
1333 However, if the real host name was known, it was omitting the HELO data
1334 if it matched the actual IP address. This has been tidied up so that it
1335 doesn't show the same IP address twice.
1337 PH/07 When both +timestamp and +memory debugging was on, the value given by
1338 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1339 called by the debug printing, thereby overwriting the timestamp buffer.
1340 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1342 PH/08 When the original message was included in an autoreply transport, it
1343 always said "this is a copy of the message, including all the headers",
1344 even if body_only or headers_only was set. It now gives an appropriate
1347 PH/09 Applied a patch from the Sieve maintainer which:
1349 o fixes some comments
1350 o adds the (disabled) notify extension core
1351 o adds some debug output for the result of if/elsif tests
1352 o points to the current vacation draft in the documentation
1353 and documents the missing references header update
1357 o fixes a bug in processing the envelope test (when testing
1358 multiple envelope elements, the last element determinted the
1361 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1362 Electronic Mail") by including:
1364 Auto-submitted: auto-generated
1366 in the messages that it generates (bounce messages and others, such as
1367 warnings). In the case of bounce messages for non-SMTP mesages, there was
1368 also a typo: it was using "Auto_submitted" (underscore instead of
1369 hyphen). Since every message generated by Exim is necessarily in response
1370 to another message, thes have all been changed to:
1372 Auto-Submitted: auto-replied
1374 in accordance with these statements in the RFC:
1376 The auto-replied keyword:
1378 - SHOULD be used on messages sent in direct response to another
1379 message by an automatic process,
1381 - MUST NOT be used on manually-generated messages,
1383 - MAY be used on Delivery Status Notifications (DSNs) and Message
1384 Disposition Notifications (MDNs),
1386 - MUST NOT be used on messages generated by automatic or periodic
1387 processes, except for messages which are automatic responses to
1390 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1391 to the default Received: header definition.
1393 PH/12 Added log selector acl_warn_skipped (default on).
1395 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1396 variables because (a) they are in the wrong storage pool and (b) even if
1397 they were copied, it wouldn't work properly because of the caching.
1399 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1400 checking when decoding. Apparently there are clients that generate
1401 overlong encoded strings. Why am I not surprised?
1403 PH/15 If the first argument of "${if match_address" was not empty, but did not
1404 contain an "@" character, Exim crashed. Now it writes a panic log message
1405 and treats the condition as false.
1407 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1409 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1410 extension "envelope-auth". The code is finished and in agreement with
1411 other implementations, but there is no documentation so far and in fact,
1412 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1413 not changing the active code.
1415 Print executed "if" and "elsif" statements when debugging is used. This
1416 helps a great deal to understand what a filter does.
1418 Document more things not specified clearly in RFC3028. I had all this
1419 sorted out, when out of a sudden new issues came to my mind. Oops."
1421 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1424 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1425 canonical form (as documented). However, after a host name lookup from
1426 the IP address, check_host() was doing a simple string comparison with
1427 addresses acquired from the DNS when checking that the found name did
1428 have the original IP as one of its addresses. Since any found IPv6
1429 addresses are likely to be in abbreviated form, the comparison could
1430 fail. Luckily, there already exists a function for doing the comparison
1431 by converting both addresses to binary, so now that is used instead of
1432 the text comparison.
1434 PH/20 There was another similar case to PH/19, when a complete host name was
1435 given in a host list; looking up its IP address could give an abbreviated
1436 form, whereas the current host's name might or might not be abbreviated.
1437 The same fix has been applied.
1443 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1444 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1447 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1448 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1450 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1452 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1453 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1454 Also, for "Auto-Submitted", treat anything other than "no" as
1455 non-personal, in accordance with RFC 3834. (Previously it treated
1456 anything starting "auto-" as non-personal.)
1458 TF/01 The control=submission/name=... option had a problem with syntax
1459 errors if the name included a slash character. The /name= option
1460 now slurps the rest of the string, so it can include any characters
1461 but it must come last in the list of options (after /sender_retain
1464 PH/05 Some modifications to the interface to the fake nameserver for the new
1472 TK/01 Added the "success_on_redirect" address verification option. See
1473 NewStuff for rationale and an example.
1475 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1477 PH/02 Patch to exigrep to allow it to work on syslog lines.
1479 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1480 fread() to skip over the body file's header line, because in Cygwin the
1481 header line is locked and is inaccessible.
1483 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1484 co-exist for some time) to make it clear that it is the Exim ID that is
1485 referenced, not the Message-ID: header line.
1487 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1488 string_format() function, because snprintf() does not exist on all
1491 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1492 ${stat: expansion item.
1494 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1495 protocol synchronization error", to keep the pedants happy.
1497 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1498 well as for IRIX systems, when gcc is being used. See the host.c source
1501 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1503 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1506 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1507 transport and to the smtp transport in LMTP mode.
1509 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1511 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1512 run a filter in a subprocess. This could lead to confusion in subsequent
1513 lookups in the parent process. There should also be a search_tidyup() at
1514 the end of the subprocess.
1516 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1517 only if the host matched helo_try_verify_hosts, which caused the
1518 verification to occur when the EHLO/HELO command was issued. The ACL just
1519 tested the remembered result. Now, if a previous verification attempt has
1520 not happened, "verify = helo" does it there and then.
1522 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1524 TK/03 Fix log output including CR from clamd.
1526 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1527 exist provoked a memory error which could cause a segfault.
1529 PH/15 Installed PCRE 6.2
1531 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1533 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1534 of the problem. Specifically, suggested +O2 rather than +O1 for the
1537 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1539 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1540 record, the port was not correctly logged when the outgoing_port log
1541 selector was set (it logged the transort's default port).
1543 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1544 fallback_hosts, and "hosts" in the smtp transport.
1546 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1547 host errors such as "Connection refused".
1549 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1550 authentication with radiusclient 0.4.9:
1552 - Error returned from rc_read_config was caught wrongly
1553 - Username/password not passed on to radius server due to wrong length.
1555 The presumption is that some radiusclient API changes for 4.51/PH/17
1556 were not taken care of correctly. The code is still untested by me (my
1557 Linux distribution still has 0.3.2 of radiusclient), but it was
1558 contributed by a Radius user.
1560 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1561 expanding the "port" option of the smtp transport.
1563 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1564 while reading a MIME header. Thanks to Tom Hughes for a patch.
1566 PH/24 Include config.h inside local_scan.h so that configuration settings are
1569 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1570 that in an ACL for RCPT (for example), you can examine exactly what was
1573 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1574 commands, but it was not correctly comparing the address with the actual
1575 client host address. Thus, it would show the EHLO address in Received:
1576 header lines when this was not necessary.
1578 PH/27 Added the % operator to ${eval:}.
1580 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1581 it should be ignoring failures (because with -C, for example, it has lost
1582 privilege). It wasn't ignoring creation failures other than "already
1585 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1588 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1589 because an input line was too long, either on its own, or by virtue of
1590 too many continuations, the temporary file was not being removed, and the
1591 return code was incorrect.
1593 PH/31 Missing "BOOL" in function definition in filtertest.c.
1595 PH/32 Applied Sieve patches from the maintainer.
1597 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1599 PH/33 Added "verify = not_blind".
1601 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1602 Local/Makefile (with some defaults set). These are used in built scripts
1603 such as exicyclog, but they have never been used in the exim_install
1604 script (though there are many overriding facilities there). I have
1605 arranged that the exim_install script now takes note of these two
1608 PH/35 Installed configuration files for Dragonfly.
1610 PH/36 When a locally submitted message by a trusted user did not contain a
1611 From: header, and the sender address was obtained from -f or from an SMTP
1612 MAIL command, and the trusted user did not use -F to supply a sender
1613 name, $originator_name was incorrectly used when constructing a From:
1614 header. Furthermore, $originator_name was used for submission mode
1615 messages from external hosts without From: headers in a similar way,
1616 which is clearly wrong.
1618 PH/37 Added control=suppress_local_fixups.
1620 PH/38 When log_selector = +received_sender was set, and the addition of the
1621 sender made the log line's construction buffer exactly full, or one byte
1622 less than full, an overflow happened when the terminating "\n" was
1625 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1626 when the result of a list match is failure because a DNS lookup failed.
1628 PH/40 RM_COMMAND is now used in the building process.
1630 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1631 the "build-* directories that it finds.
1633 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1634 address in a domain literal was a prefix of an interface address.
1636 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1637 when verifying a sender address, unless rewrite_headers is false.
1639 PH/44 Wrote a long comment about why errors_to addresses are verified as
1640 recipients, not senders.
1642 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1643 the ratelimit ACL was added.
1645 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1647 PH/46 Added extra information about PostgreSQL errors to the error string.
1649 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1650 avoiding the need to install special zones in a real server. This is
1651 backwards compatible; if it can't find the fake resolver, it drops back.
1652 Thus, both old and new test suites can be run.
1654 TF/02 Added util/ratelimit.pl
1656 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1657 clock is set back in time.
1659 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1660 Candler <B.Candler@pobox.com>.
1662 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1663 OK for addresses that are the result of redirections.
1665 PH/48 A number of further additions for the benefit of the new test suite,
1666 including a fake gethostbyname() that interfaces to the fake DNS resolver
1669 TF/06 The fix for widen_domains has also been applied to qualify_single and
1670 search_parents which are the other dnslookup options that can cause
1673 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1676 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1678 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1679 service" if it encountered an MX record with an empty target hostname.
1680 The message is now "an MX or SRV record indicated no SMTP service".
1682 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1683 if the DNS of the sending site is misconfigured. This is quite a
1684 common situation. This change restores the behaviour of treating a
1685 helo verification defer as a failure.
1687 PH/51 If self=fail was set on a router, the bounce message did not include the
1688 actual error message.
1694 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1696 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1697 command itself ended in error, the underlying message about the transport
1698 filter timeout was being overwritten with the pipe command error. Now the
1699 underlying error message should be appended to the second error message.
1701 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1703 PH/02 Reduce the amount of output that "make" produces by default. Full output
1704 can still be requested.
1706 PH/03 The warning log line about a condition test deferring for a "warn" verb
1707 was being output only once per connection, rather than after each
1708 occurrence (because it was using the same function as for successful
1709 "warn" verbs). This seems wrong, so I have changed it.
1711 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1712 it should not have, which might have caused a crash in the right
1713 circumstances, but probably never did.
1715 PH/04 Installed a modified version of Tony Finch's patch to make submission
1716 mode fix the return path as well as the Sender: header line, and to
1717 add a /name= option so that you can make the user's friendly name appear
1720 TF/03 Added the control = fakedefer ACL modifier.
1722 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1723 Mark Lowes for thorough testing.
1725 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1727 TK/03 Merged latest SRS patch from Miles Wilton.
1729 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1730 with the definition in sysexits.h (which is #included earlier).
1731 Fortunately, Exim does not actually use EX_OK. The code used to try to
1732 preserve the sysexits.h value, by assumimg that macro definitions were
1733 scanned for macro replacements. I have been disabused of this notion,
1734 so now the code just undefines EX_OK before #including unistd.h.
1736 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1737 in the smtp transport. When a block could not be written in a single
1738 write() function, the timeout was being re-applied to each part-write.
1739 This seems wrong - if the receiver was accepting one byte at a time it
1740 would take for ever. The timeout is now adjusted when this happens. It
1741 doesn't have to be particularly precise.
1743 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1744 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1746 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1747 without a domain if the check to <postmaster@domain> fails.
1749 SC/01 Eximstats: added -xls and the ability to specify output files
1750 (patch written by Frank Heydlauf).
1752 SC/02 Eximstats: use FileHandles for outputing results.
1754 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1756 SC/04 Eximstats: fixed display of large numbers with -nvr option
1758 SC/05 Eximstats: fixed merging of reports with empty tables.
1760 SC/06 Eximstats: added the -include_original_destination flag
1762 SC/07 Eximstats: removed tabs and trailing whitespace.
1764 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1766 TK/06 MBOX spool code: Add real "From " MBOX separator line
1767 so the .eml file is really in mbox format (even though
1768 most programs do not really care). Patch from Alex Miller.
1770 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1771 The latter is generated from $received_to and is only set if the
1772 message has one envelope recipient. SA can use these headers,
1773 obviously out-of-the-box. Patch from Alex Miller.
1775 PH/08 The ${def test on a variable was returning false if the variable's
1776 value was "0", contrary to what the specification has always said!
1777 The result should be true unless the variable is empty.
1779 PH/09 The syntax error of a character other than { following "${if
1780 def:variable_name" (after optional whitespace) was not being diagnosed.
1781 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1782 accidental colon was present, for example, could give incorrect results.
1784 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1785 result is used (not including appendfile, where other changes are about
1788 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1789 This involved changing a lot of size variables from int to off_t. It
1790 should work with maildirs and everything.
1792 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1793 spamd dying while we are connected to it.
1795 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1798 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1799 function for BDB 4.3.
1801 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1803 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1804 the sources for PCRE have been re-arranged and more files are now
1807 PH/15 The code I had for printing potentially long long variables in PH/11
1808 above was not the best (it lost precision). The length of off_t variables
1809 is now inspected at build time, and an appropriate printing format (%ld
1810 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1811 to be "long long int" or "long int". This is needed for the internal
1812 formatting function string_vformat().
1814 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1815 the configuration file to be ":syslog", then the script "guesses" where
1816 the logs files are, rather than using the compiled in default. In our
1817 case the guess is not the same as the compiled default, so the script
1818 suddenly stopped working when I started to use syslog. The patch checks
1819 to see if log_file_path is "". If so, it attempts to read it from exim
1820 with no configuration file to get the compiled in version, before it
1821 falls back to the previous guessing code."
1823 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1824 implementing BATV in an Exim configuration. See NewStuff for the gory
1827 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1828 Makefile that are specific to HP-UX.
1830 PH/18 If the "use_postmaster" option was set for a recipient callout together
1831 with the "random" option, the postmaster address was used as the MAIL
1832 FROM address for the random test, but not for the subsequent recipient
1833 test. It is now used for both.
1835 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1836 patch removes a few documentation additions to RFC 3028, because the
1837 latest draft now contains them. It adds the new en;ascii-case comparator
1838 and a new error check for 8bit text in MIME parts. Comparator and
1839 require names are now matched exactly. I enabled the subaddress
1840 extension, but it is not well tested yet (read: it works for me)."
1842 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1843 rework some of the code of TK/09 above to avoid the hardwired use of
1844 "%lld" and "long long". Replaced the call to snprintf() with a call to
1847 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1848 records point to non-existent hosts", "retry timeout exceeded", and
1849 "retry time not reached for any host after a long failure period".
1851 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1852 experimental DomainKeys support:
1854 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1855 (2) On an error such as an illegally used "control", the wrong name for
1856 the control was given.
1858 These problems did NOT occur unless DomainKeys support was compiled.
1860 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1862 PH/24 Added ${if match_ip condition.
1864 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1865 databases so that it will be absolutely obvious if a crash occurs in the
1866 DB library. This is a regular occurrence (often caused by mis-matched
1869 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1870 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1871 picked up on a user's system that detects such things. There doesn't seem
1872 to be a gcc warning option for this - only an attribute that has to be
1873 put on the function's prototype. It seems that in Fedora Core 4 they have
1874 set this on a number of new functions. No doubt there will be more in due
1877 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1878 specify a transport. However, if an address that was verified by such a
1879 router was the subject of a callout, Exim crashed because it tried to
1880 read the rcpt_include_affixes from the non-existent transport. Now it
1881 just assumes that the setting of that option is false. This bug was
1882 introduced by 4.51/PH/31.
1884 PH/28 Changed -d+all to exclude +memory, because that information is very
1885 rarely of interest, but it makes the output a lot bigger. People tend to
1886 do -d+all out of habit.
1888 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1889 code in os-type was giving problems when libc.so lives in lib64, like on
1892 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1893 aren't the modern standard, and it seems that some systems' include files
1894 don't always have them. Exim was already checking for some of the newer
1895 ones like T_AAAA, and defining it itself. I've added checks for all the
1896 record types that Exim uses.
1898 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1899 not automatically generating a new one, as it is supposed to. This
1900 prevented TLS from working. If the file did exist, but contained invalid
1901 data, a new version was generated, as expected. It was only the case of a
1902 non-existent file that was broken.
1904 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1905 with a change in libdomainkeys > 0.64.
1907 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1908 from DNS. If the selector record carries the flag, it now has
1909 precedence over the domain-wide flag.
1911 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1913 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1914 the use of a port name that isn't defined in /etc/services) occurred, the
1915 message was deferred as in a normal delivery, and thus remained on the
1916 spool, instead of being failed because of the mua_wrapper setting. This
1917 is now fixed, and I tidied up some of the mua_wrapper messages at the
1920 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1921 the messages in a hash of arrays rather than using individual hashes.
1922 This is a bit cleaner and results in dramatic memory savings, albeit
1923 at a slight CPU cost.
1925 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1926 as requested by Marc Sherman.
1928 SC/10 Eximstats: added histograms for user specified patterns as requested
1931 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1933 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1934 fopen() in the content-scanning modules that did not already have it.
1940 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1941 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1943 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1945 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1947 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1948 to test Sieve filters that use "vacation".
1950 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1951 that changes the way the GnuTLS parameters are stored in the cache file.
1952 The new format can be generated externally. For backward compatibility,
1953 if the data in the cache doesn't make sense, Exim assumes it has read an
1954 old-format file, and it generates new data and writes a new file. This
1955 means that you can't go back to an older release without removing the
1958 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1959 work if there are any delivery delays because "one_time" forces the
1960 parent to be marked "delivered", so its unseen clone is never tried
1961 again. For this reason, Exim now forbids the simultaneous setting of
1964 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1965 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1966 entry if you want to know the details. The fix, however, neglected to
1967 consider the case when local delivery batching is involved. The test for
1968 "previously delivered" was not happening when checking to see if an
1969 address could be batched with a previous (undelivered) one; under
1970 certain circumstances this could lead to multiple deliveries to the same
1973 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1974 in its include files, and this causes problems building Exim.
1976 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1977 header_syntax) but Exim was just ignoring anything given after a slash.
1978 In particular, this caused confusion with an attempt to use "verify =
1979 reverse_host_lookup/defer_ok". An error is now given when options are
1980 supplied for verify items that do not have them. (Maybe reverse_host_
1981 lookup should have a defer_ok option, but that's a different point.)
1983 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1984 defined by RFC 821) to 2048, because there were problems with some AUTH
1985 commands, and RFC 1869 says the size should be increased for extended
1986 SMTP commands that take arguments.
1988 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1991 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1992 "unknown" error; now it says that the functionality isn't in the binary.
1994 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1995 an address' error message when a string expansion fails (syntax or
1996 whatever). Otherwise the password may appear in the log. Following change
1997 PH/42 below, there is no longer a chance of it appearing in a bounce
2000 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2002 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2003 of its addresses was ever tried. (Bugzilla bug #2.)
2005 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2006 the result incorrectly in the debug output. (It correctly added a newline
2007 to what was transported.)
2009 TF/01 Added $received_time.
2011 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2012 commented out examples of how to interface to a virus scanner and to
2013 SpamAssassin. Also added commented examples of av_scanner and
2014 spamd_address settings.
2016 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2017 and controls are allowed in which ACLs. There were a couple of minor
2018 errors. Some of the entries in the conditions table (which is a table of
2019 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2020 negation of where the condition IS allowed.
2022 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2024 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2025 header file does not have a version number, so I've had to invent a new
2026 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2027 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2028 radiusclient), but it was contributed by a Radius user.
2030 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2031 files or directories, for OpenSSL.
2033 PH/19 When an Exim process that is running as root has to create an Exim log
2034 file, it does so in a subprocess that runs as exim:exim so as to get the
2035 ownership right at creation (otherwise, other Exim processes might see
2036 the file with the wrong ownership). There was no test for failure of this
2037 fork() call, which would lead to the process getting stuck as it waited
2038 for a non-existent subprocess. Forks do occasionally fail when resources
2039 run out. I reviewed all the other calls to fork(); they all seem to check
2042 PH/20 When checking for unexpected SMTP input at connect time (before writing
2043 the banner), Exim was not dealing correctly with a non-positive return
2044 from the read() function. If the client had disconnected by this time,
2045 the result was a log entry for a synchronization error with an empty
2046 string after "input=" when read() returned zero. If read() returned -1
2047 (an event I could not check), uninitialized data bytes were printed.
2048 There were reports of junk text (parts of files, etc) appearing after
2051 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2053 PH/22 Added support for macro redefinition, and (re)definition in between
2054 driver and ACL definitions.
2056 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2057 forgetting to use the resulting value; it was using the unexpanded value.
2059 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2060 hadn't been configured. The fix is from Juergen Kreileder, who
2061 understands it better than I do:
2063 "Here's what I see happening with three configured cyrus_sasl
2064 authenticators configured (plain, login, cram-md5):
2066 On startup auth_cyrus_sasl_init() gets called for each of these.
2067 This means three calls to sasl_listmech() without a specified mech_list.
2068 => SASL tests which mechs of all available mechs actually work
2069 => three warnings about OTP not working
2070 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2072 With the patch, sasl_listmech() also gets called three times. But now
2073 SASL's mech_list option is set to the server_mech specified in the the
2074 authenticator. Or in other words, the answer from sasl_listmech()
2075 gets limited to just the mech you're testing for (which is different
2077 => the return list contains just 'plain' or 'login', 'cram-md5' or
2078 nothing depending on the value of ob->server_mech.
2080 I've just tested the patch: Authentication still works fine,
2081 unavailable mechs specified in the exim configuration are still
2082 caught, and the auth.log warnings about OTP are gone."
2084 PH/25 When debugging is enabled, the contents of the command line are added
2085 to the debugging output, even when log_selector=+arguments is not
2088 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2089 answer is "GNU", and only if the return is "GNU/something" is the answer
2092 PH/27 $acl_verify_message is now set immediately after the failure of a
2093 verification in an ACL, and so is available in subsequent modifiers. In
2094 particular, the message can be preserved by coding like this:
2096 warn !verify = sender
2097 set acl_m0 = $acl_verify_message
2099 Previously, $acl_verify_message was set only while expanding "message"
2100 and "log_message" when a very denied access.
2102 PH/28 Modified OS/os.c-Linux with
2104 -#ifndef OS_LOAD_AVERAGE
2105 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2107 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2108 nomenclature these days.)
2110 PH/29 Installed patch from the Sieve maintainer that adds the options
2111 sieve_useraddress and sieve_subaddress to the redirect router.
2113 PH/30 In these circumstances:
2114 . Two addresses routed to the same list of hosts;
2115 . First host does not offer TLS;
2116 . First host accepts first address;
2117 . First host gives temporary error to second address;
2118 . Second host offers TLS and a TLS session is established;
2119 . Second host accepts second address.
2120 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2121 and peerdn, if requested) that were in fact used only for the second
2124 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2125 attention to any local part prefix or suffix that was matched by the
2126 router that accepted the address. It now behaves in the same way as it
2127 does for delivery: the affixes are removed from the local part unless
2128 rcpt_include_affixes is set on the transport.
2130 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2131 timeout during the DATA phase of an incoming message.
2133 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2134 applied a patch sanctioned by the Sieve maintainer.
2136 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2137 the uid or gid is negative. A case of a negative gid caused this to be
2138 noticed. The fix allows for either to be negative.
2140 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2141 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2144 PH/36 The change PH/12 above was broken. Fixed it.
2146 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2147 the grounds that routing the same address twice would always produce the
2148 same answer. This might have been true once, but it is certainly no
2149 longer true now. Routing a child address may depend on the previous
2150 routing that produced that child. Some complicated redirection strategies
2151 went wrong when messages had multiple recipients, and made Exim's
2152 behaviour dependent on the order in which the addresses were given.
2154 I have moved the duplicate checking until after the routing is complete.
2155 Exim scans the addresses that are assigned to local and remote
2156 transports, and removes any duplicates. This means that more work will be
2157 done, as duplicates will always all be routed, but duplicates are
2158 presumably rare, so I don't expect this is of any significance.
2160 For deliveries to pipes, files, and autoreplies, the duplicate checking
2161 still happens during the routing process, since they are not going to be
2164 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2165 It corrects a timeout issue with spamd. This is Ian's comment: "The
2166 background is that sometimes spamd either never reads data from a
2167 connection it has accepted, or it never writes response data. The exiscan
2168 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2169 blindly assumes that writes won't block so it may never time out."
2171 PH/39 Allow G after quota size as well as K and M.
2173 PH/40 The value set for $authenticated_id in an authenticator may not contain
2174 binary zeroes or newlines because the value is written to log lines and
2175 to spool files. There was no check on this. Now the value is run through
2176 the string_printing() function so that such characters are converted to
2177 printable escape sequences.
2179 PH/41 $message_linecount is a new variable that contains the total number of
2180 lines in the message. Compare $body_linecount, which is the count for the
2183 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2184 bounce and delay warning messages, except in certain special cases, which
2187 (a) An SMTP error message from a remote host;
2188 (b) A message specified in a :fail: redirection;
2189 (c) A message specified in a "fail" command in a system filter;
2190 (d) A message specified in a FAIL return from the queryprogram router;
2191 (e) A message specified by the cannot_route_message router option.
2193 In these cases only, Exim does include the error details in bounce and
2194 warning messages. There are also a few cases where bland messages such
2195 as "unrouteable address" or "local delivery error" are given.
2197 PH/43 $value is now also set for the "else" part of a ${run expansion.
2199 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2200 being worked on, but at least Exim now implements the latest version to
2203 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2204 process to complete was treated as a delivery failure, a timeout while
2205 writing the message to the pipe was logged, but erroneously treated as a
2206 successful delivery. Such timeouts include transport filter timeouts. For
2207 consistency with the overall process timeout, these timeouts are now
2208 treated as errors, giving rise to delivery failures by default. However,
2209 there is now a new Boolean option for the pipe transport called
2210 timeout_defer, which, if set TRUE, converts the failures into defers for
2211 both kinds of timeout. A transport filter timeout is now identified in
2214 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2215 systems where "make" and "gmake" are different, calling "gmake" at top
2216 level broke things. I've arranged for the value of $(MAKE) to be passed
2217 from the Makefile to this script so that it can call the same version of
2221 A note about Exim versions 4.44 and 4.50
2222 ----------------------------------------
2224 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2225 changes of various kinds. As a consequence, a big documentation update was
2226 needed. This delayed the release for rather longer than seemed good, especially
2227 in the light of a couple of (minor) security issues. Therefore, the changes
2228 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2229 release. So 4.44 and 4.50 are in effect two different branches that both start
2232 I have left the 4.50 change log unchanged; it contains all the changes since
2233 4.43. The change log for 4.44 is below; many of its items are identical to
2234 those for 4.50. This seems to be the most sensible way to preserve the
2235 historical information.
2241 1. Minor wording change to the doc/README.SIEVE file.
2243 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2244 computation of the current number of files was incorrect.
2246 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2247 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2250 4. Give more explanation in the error message when the command for a transport
2251 filter fails to execute.
2253 5. There are several places where Exim runs a non-Exim command in a
2254 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2255 was being done only for the command run by the queryprogram router. It is
2256 now done for all such subprocesses. The other cases are: ${run, transport
2257 filters, and the commands run by the lmtp and pipe transports.
2259 6. Added CONFIGURE_GROUP build-time option.
2261 7. Some older OS have a limit of 256 on the maximum number of file
2262 descriptors. Exim was using setrlimit() to set 1000 as a large value
2263 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2264 systems. I've change it so that if it can't get 1000, it tries for 256.
2266 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2267 was an oversight, and furthermore, ever since the addition of extra
2268 controls (e.g. 4.43/32), the checks on when to allow different forms of
2269 "control" were broken. There should now be diagnostics for all cases when a
2270 control that does not make sense is encountered.
2272 9. Added the /retain_sender option to "control=submission".
2274 10. $recipients is now available in the predata ACL (oversight).
2276 11. Tidy the search cache before the fork to do a delivery from a message
2277 received from the command line. Otherwise the child will trigger a lookup
2278 failure and thereby defer the delivery if it tries to use (for example) a
2279 cached ldap connection that the parent has called unbind on.
2281 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2282 of $address_data from the recipient verification was clobbered by the
2283 sender verification.
2285 13. The value of address_data from a sender verification is now available in
2286 $sender_address_data in subsequent conditions in the ACL statement.
2288 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2290 15. Added a new option "connect=<time>" to callout options, to set a different
2293 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2294 was its contents. (It was OK if the option was not defined at all.)
2296 17. A "Completed" log line is now written for messages that are removed from
2297 the spool by the -Mrm option.
2299 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2300 information about exactly what failed.
2302 19. Added -dd to debug only the daemon process.
2304 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2305 handles timeouts, both on the server side and network timeouts. Renamed the
2306 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2308 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2309 It is now set to "smtps".
2311 22. $host_address is now set to the target address during the checking of
2312 ignore_target_hosts.
2314 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2315 being passed; this would have caused $sender_host_name to have been used if
2316 matching the list had actually called for a host name (not very likely,
2317 since this list is usually IP addresses). A host name is now passed as
2320 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2321 code that specifies a non-restarting handler (typically sigaction() in
2322 modern systems) in an attempt to fix a rare and obscure crash bug.
2324 25. Narrowed the window for a race in the daemon that could cause it to ignore
2325 SIGCHLD signals. This is not a major problem, because they are used only to
2326 wake it up if nothing else does.
2328 26. A malformed maildirsize file could cause Exim to calculate negative values
2329 for the mailbox size or file count. Odd effects could occur as a result.
2330 The maildirsize information is now recalculated if the size or filecount
2333 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2334 support for a long time. Removed HAVE_SYS_VFS_H.
2336 28. Installed the latest version of exipick from John Jetmore.
2338 29. In an address list, if the pattern was not a regular expression, an empty
2339 subject address (from a bounce message) matched only if the pattern was an
2340 empty string. Non-empty patterns were not even tested. This was the wrong
2341 because it is perfectly reasonable to use an empty address as part of a
2342 database query. An empty address is now tested by patterns that are
2343 lookups. However, all the other forms of pattern expect the subject to
2344 contain a local part and a domain, and therefore, for them, an empty
2345 address still always fails if the pattern is not itself empty.
2347 30. Exim went into a mad DNS loop when attempting to do a callout where the
2348 host was specified on an smtp transport, and looking it up yielded more
2349 than one IP address.
2351 31. Re-factored the code for checking spool and log partition space into a
2352 function that finds that data and another that does the check. The former
2353 is then used to implement four new variables: $spool_space, $log_space,
2354 $spool_inodes, and $log_inodes.
2356 32. The RFC2047 encoding function was originally intended for short strings
2357 such as real names; it was not keeping to the 75-character limit for
2358 encoded words that the RFC imposes. It now respects the limit, and
2359 generates multiple encoded words if necessary. To be on the safe side, I
2360 have increased the buffer size for the ${rfc2047: expansion operator from
2363 33. It is now permitted to omit both strings after an "if" condition; if the
2364 condition is true, the result is "true". As before, when the second string
2365 is omitted, a false condition yields an empty string. This makes it less
2366 cumbersome to write custom ACL and router conditions.
2368 34. Failure to deliver a bounce message always caused it to be frozen, even if
2369 there was an errors_to setting on the router. The errors_to setting is now
2372 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2373 canonical form (fully expanded) before being placed in
2374 $sender_host_address.
2376 36. The table in the code that translates DNS record types into text (T_A to
2377 "A" for instance) was missing entries for NS and CNAME. It is just possible
2378 that this could have caused confusion if both these types were looked up
2379 for the same domain, because the text type is used as part of Exim's
2380 per-process caching. But the chance of anyone hitting this buglet seems
2383 37. The dnsdb lookup has been extended in a number of ways.
2385 (1) There is a new type, "zns", which walks up the domain tree until it
2386 finds some nameserver records. It should be used with care.
2388 (2) There is a new type, "mxh", which is like "mx" except that it returns
2389 just the host names, not the priorities.
2391 (3) It is now possible to give a list of domains (or IP addresses) to be
2392 looked up. The behaviour when one of the lookups defers can be
2393 controlled by a keyword.
2395 (4) It is now possible to specify the separator character for use when
2396 multiple records are returned.
2398 38. The dnslists ACL condition has been extended: it is now possible to supply
2399 a list of IP addresses and/or domains to be looked up in a particular DNS
2402 39. Added log_selector=+queue_time_overall.
2404 40. When running the queue in the test harness, wait just a tad after forking a
2405 delivery process, to get repeatability of debugging output.
2407 41. Include certificate and key file names in error message when GnuTLS fails
2408 to set them up, because the GnuTLS error message doesn't include the name
2409 of the failing file when there is a problem reading it.
2411 42. Allow both -bf and -bF in the same test run.
2413 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2415 44. The "Exiscan patch" is now merged into the mainline Exim source.
2417 45. Sometimes the final signoff response after QUIT could fail to get
2418 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2419 < 0 before doing a fflush(). This bug looks as though it goes back to the
2420 introduction of TLS in release 3.20, but "sometimes" must have been rare
2421 because the tests only now provoked it.
2423 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2424 (this can affect the format of dates).
2426 47. exim_tidydb, when checking for the continued existence of a message for
2427 which it has found a message-specific retry record, was not finding
2428 messages that were in split spool directories. Consequently, it was
2429 deleting retry records that should have stayed in existence.
2431 48. Steve fixed some bugs in eximstats.
2433 49. The SPA authentication driver was not abandoning authentication and moving
2434 on to the next authenticator when an expansion was forced to fail,
2435 contradicting the general specification for all authenticators. Instead it
2436 was generating a temporary error. It now behaves as specified.
2438 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2439 (the order specifies the preference for clients). The order is now AES256,
2440 AES128, 3DES, ARCFOUR128.
2442 51. Small patch to Sieve code - explicitly set From: when generating an
2445 52. Exim crashed if a remote delivery caused a very long error message to be
2446 recorded - for instance if somebody sent an entire SpamAssassin report back
2447 as a large number of 550 error lines. This bug was coincidentally fixed by
2448 increasing the size of one of Exim's internal buffers (big_buffer) that
2449 happened as part of the Exiscan merge. However, to be on the safe side, I
2450 have made the code more robust (and fixed the comments that describe what
2453 53. Now that there can be additional text after "Completed" in log lines (if
2454 the queue_time_overall log selector is set), a one-byte patch to exigrep
2455 was needed to allow it to recognize "Completed" as not the last thing in
2458 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2459 patch that reportedly fixes this has been added. I am not expert enough to
2460 create a test for it. This is what the patch creator wrote:
2462 "I found a little strange behaviour of ldap code when working with
2463 Windows 2003 AD Domain, where users was placed in more than one
2464 Organization Units. When I tried to give exim partial DN, the exit code
2465 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2466 But simultaneously result of request was absolutely normal ldap result,
2467 so I produce this patch..."
2469 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2470 REFERENCE, so I have modified the code to exclude the patch when that macro
2473 55. Some experimental protocols are using DNS PTR records for new purposes. The
2474 keys for these records are domain names, not reversed IP addresses. The
2475 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2476 leaves it alone. Component reversal etc. now happens only for IP addresses.
2479 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2481 57. Double the size of the debug message buffer (to 2048) so that more of very
2482 long debug lines gets shown.
2484 58. The exicyclog utility now does better if the number of log files to keep
2485 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2487 59. Two changes related to the smtp_active_hostname option:
2489 (1) $smtp_active_hostname is now available as a variable.
2490 (2) The default for smtp_banner uses $smtp_active_hostname instead
2491 of $primary_hostname.
2493 60. The host_aton() function is supposed to be passed a string that is known
2494 to be a valid IP address. However, in the case of IPv6 addresses, it was
2495 not checking this. This is a hostage to fortune. Exim now panics and dies
2496 if the condition is not met. A case was found where this could be provoked
2497 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2498 components; fortuitously, this particular loophole had already been fixed
2499 by change 4.50/55 above.
2501 If there are any other similar loopholes, the new check in host_aton()
2502 itself should stop them being exploited. The report I received stated that
2503 data on the command line could provoke the exploit when Exim was running as
2504 exim, but did not say which command line option was involved. All I could
2505 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2506 running as the user.
2509 61. There was a buffer overflow vulnerability in the SPA authentication code
2510 (which came originally from the Samba project). I have added a test to the
2511 spa_base64_to_bits() function which I hope fixes it.
2514 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2515 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2517 63. The daemon start-up calls getloadavg() while still root for those OS that
2518 need the first call to be done as root, but it missed one case: when
2519 deliver_queue_load_max is set with deliver_drop_privilege. This is
2520 necessary for the benefit of the queue runner, because there is no re-exec
2521 when deliver_drop_privilege is set.
2523 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2524 This has been fixed.
2526 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2527 was in use, was not putting the data itself into the right store pool;
2528 consequently, it could be overwritten for a subsequent message in the same
2529 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2532 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2534 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2535 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2536 the function were treating the return as a boolean value, which happened to
2537 work because 0=false and not-0=true, but is not correct code.
2539 68. The host_aton() function was not handling scoped IPv6 addresses (those
2540 with, for example, "%eth0" on the end) correctly.
2542 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2543 negated items (that is, ~something) in unsigned ints. Some compilers
2544 apparently mutter when there is no cast.
2546 70. If an address verification called from an ACL failed, and did not produce a
2547 user-specific message (i.e. there was only a "system" message), nothing was
2548 put in $acl_verify_message. In this situation, it now puts the system
2551 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2552 session; change 4.31/43 added the unwanted input to the log line - except
2553 that it did not do this in the start of session case. It now does.
2555 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2556 This is wrong and can cause the other end to generate a synchronization
2557 error if it is another Exim or anything else that does the synchronization
2558 check. A QUIT command is no longer sent after a timeout.
2560 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2561 during host lookups.
2563 74. The defer_ok option of callout verification was not working if it was used
2564 when verifying addresses in header lines, that is, for this case:
2566 verify = header_sender/callout=defer_ok
2568 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2569 those file descriptors could be used for SMTP connections. If anything
2570 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2571 could be sent to the SMTP client, causing chaos. The daemon now opens
2572 stdin, stdout, and stderr to /dev/null when it puts itself into the
2575 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2576 log by default. The user can override this with suitable Perl magic.
2578 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2579 the log message when discard triggers, was not working for the DATA ACL or
2580 for the non-SMTP ACL.
2582 78. Error message wording change in sieve.c.
2584 79. If smtp_accept_max_per_host was set, the number of connections could be
2585 restricted to fewer than expected, because the daemon was trying to set up
2586 a new connection before checking whether the processes handling previous
2587 connections had finished. The check for completed processes is now done
2588 earlier. On busy systems, this bug wouldn't be noticed because something
2589 else would have woken the daemon, and it would have reaped the completed
2592 80. If a message was submitted locally by a user whose login name contained one
2593 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2594 It caused a spool format error. I have fixed the spool reading code. A
2595 related problem was that the "from" clause in the Received: line became
2596 illegal because of the space(s). It is now covered by ${quote_local_part.
2598 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2601 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2603 83. Patch from Sieve maintainer to fix unterminated string problem in
2604 "vacation" handling.
2606 84. Some minor changes to the Linux configuration files to help with other
2607 OS variants using glibc.
2609 85. One more patch for Sieve to update vacation handling to latest spec.
2612 ----------------------------------------------------
2613 See the note above about the 4.44 and 4.50 releases.
2614 ----------------------------------------------------
2620 1. Change 4.43/35 introduced a bug that caused file counts to be
2621 incorrectly computed when quota_filecount was set in an appendfile
2624 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2625 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2628 3. Give more explanation in the error message when the command for a transport
2629 filter fails to execute.
2631 4. There are several places where Exim runs a non-Exim command in a
2632 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2633 was being done only for the command run by the queryprogram router. It is
2634 now done for all such subprocesses. The other cases are: ${run, transport
2635 filters, and the commands run by the lmtp and pipe transports.
2637 5. Some older OS have a limit of 256 on the maximum number of file
2638 descriptors. Exim was using setrlimit() to set 1000 as a large value
2639 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2640 systems. I've change it so that if it can't get 1000, it tries for 256.
2642 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2643 was an oversight, and furthermore, ever since the addition of extra
2644 controls (e.g. 4.43/32), the checks on when to allow different forms of
2645 "control" were broken. There should now be diagnostics for all cases when a
2646 control that does not make sense is encountered.
2648 7. $recipients is now available in the predata ACL (oversight).
2650 8. Tidy the search cache before the fork to do a delivery from a message
2651 received from the command line. Otherwise the child will trigger a lookup
2652 failure and thereby defer the delivery if it tries to use (for example) a
2653 cached ldap connection that the parent has called unbind on.
2655 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2656 of $address_data from the recipient verification was clobbered by the
2657 sender verification.
2659 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2660 was its contents. (It was OK if the option was not defined at all.)
2662 11. A "Completed" log line is now written for messages that are removed from
2663 the spool by the -Mrm option.
2665 12. $host_address is now set to the target address during the checking of
2666 ignore_target_hosts.
2668 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2669 being passed; this would have caused $sender_host_name to have been used if
2670 matching the list had actually called for a host name (not very likely,
2671 since this list is usually IP addresses). A host name is now passed as
2674 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2675 code that specifies a non-restarting handler (typically sigaction() in
2676 modern systems) in an attempt to fix a rare and obscure crash bug.
2678 15. Narrowed the window for a race in the daemon that could cause it to ignore
2679 SIGCHLD signals. This is not a major problem, because they are used only to
2680 wake it up if nothing else does.
2682 16. A malformed maildirsize file could cause Exim to calculate negative values
2683 for the mailbox size or file count. Odd effects could occur as a result.
2684 The maildirsize information is now recalculated if the size or filecount
2687 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2688 support for a long time. Removed HAVE_SYS_VFS_H.
2690 18. Updated exipick to current release from John Jetmore.
2692 19. Allow an empty sender to be matched against a lookup in an address list.
2693 Previously the only cases considered were a regular expression, or an
2696 20. Exim went into a mad DNS lookup loop when doing a callout where the
2697 host was specified on the transport, if the DNS lookup yielded more than
2700 21. The RFC2047 encoding function was originally intended for short strings
2701 such as real names; it was not keeping to the 75-character limit for
2702 encoded words that the RFC imposes. It now respects the limit, and
2703 generates multiple encoded words if necessary. To be on the safe side, I
2704 have increased the buffer size for the ${rfc2047: expansion operator from
2707 22. Failure to deliver a bounce message always caused it to be frozen, even if
2708 there was an errors_to setting on the router. The errors_to setting is now
2711 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2712 canonical form (fully expanded) before being placed in
2713 $sender_host_address.
2715 24. Updated eximstats to version 1.33
2717 25. Include certificate and key file names in error message when GnuTLS fails
2718 to set them up, because the GnuTLS error message doesn't include the name
2719 of the failing file when there is a problem reading it.
2721 26. Expand error message when OpenSSL has problems setting up cert/key files.
2724 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2725 (this can affect the format of dates).
2727 28. exim_tidydb, when checking for the continued existence of a message for
2728 which it has found a message-specific retry record, was not finding
2729 messages that were in split spool directories. Consequently, it was
2730 deleting retry records that should have stayed in existence.
2732 29. eximstats updated to version 1.35
2733 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2734 1.35 - bugfix such that pie charts by volume are generated correctly
2736 30. The SPA authentication driver was not abandoning authentication and moving
2737 on to the next authenticator when an expansion was forced to fail,
2738 contradicting the general specification for all authenticators. Instead it
2739 was generating a temporary error. It now behaves as specified.
2741 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2742 (the order specifies the preference for clients). The order is now AES256,
2743 AES128, 3DES, ARCFOUR128.
2745 31. Small patch to Sieve code - explicitly set From: when generating an
2748 32. Exim crashed if a remote delivery caused a very long error message to be
2749 recorded - for instance if somebody sent an entire SpamAssassin report back
2750 as a large number of 550 error lines. This bug was coincidentally fixed by
2751 increasing the size of one of Exim's internal buffers (big_buffer) that
2752 happened as part of the Exiscan merge. However, to be on the safe side, I
2753 have made the code more robust (and fixed the comments that describe what
2756 33. Some experimental protocols are using DNS PTR records for new purposes. The
2757 keys for these records are domain names, not reversed IP addresses. The
2758 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2759 leaves it alone. Component reversal etc. now happens only for IP addresses.
2762 34. The host_aton() function is supposed to be passed a string that is known
2763 to be a valid IP address. However, in the case of IPv6 addresses, it was
2764 not checking this. This is a hostage to fortune. Exim now panics and dies
2765 if the condition is not met. A case was found where this could be provoked
2766 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2767 components; fortuitously, this particular loophole had already been fixed
2768 by change 4.50/55 or 4.44/33 above.
2770 If there are any other similar loopholes, the new check in host_aton()
2771 itself should stop them being exploited. The report I received stated that
2772 data on the command line could provoke the exploit when Exim was running as
2773 exim, but did not say which command line option was involved. All I could
2774 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2775 running as the user.
2778 35. There was a buffer overflow vulnerability in the SPA authentication code
2779 (which came originally from the Samba project). I have added a test to the
2780 spa_base64_to_bits() function which I hope fixes it.
2783 36. The daemon start-up calls getloadavg() while still root for those OS that
2784 need the first call to be done as root, but it missed one case: when
2785 deliver_queue_load_max is set with deliver_drop_privilege. This is
2786 necessary for the benefit of the queue runner, because there is no re-exec
2787 when deliver_drop_privilege is set.
2789 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2790 was in use, was not putting the data itself into the right store pool;
2791 consequently, it could be overwritten for a subsequent message in the same
2792 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2795 38. Sometimes the final signoff response after QUIT could fail to get
2796 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2797 < 0 before doing a fflush(). This bug looks as though it goes back to the
2798 introduction of TLS in release 3.20, but "sometimes" must have been rare
2799 because the tests only now provoked it.
2805 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2806 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2807 Now it returns an integer. A number of calls were still expecting a T/F
2808 return. Fortuitously, in all cases, the tests worked in OK situations,
2809 which is the norm. However, things would have gone wrong on any write
2810 failures on the smtp file descriptor. This function is used when sending
2811 messages over SMTP and also when doing verify callouts.
2813 2. When Exim is called to do synchronous delivery of a locally submitted
2814 message (the -odf or -odi options), it no longer closes stderr before doing
2817 3. Implemented the mua_wrapper option.
2819 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2821 5. Implemented the functions header_remove(), header_testname(),
2822 header_add_at_position(), and receive_remove_recipient(), and exported them
2825 6. If an ACL "warn" statement specified the addition of headers, Exim already
2826 inserted X-ACL-Warn: at the start if there was no header name. However, it
2827 was not making this test for the second and subsequent header lines if
2828 there were newlines in the string. This meant that an invalid header could
2829 be inserted if Exim was badly configured.
2831 7. Allow an ACL "warn" statement to add header lines at the start or after all
2832 the Received: headers, as well as at the end.
2834 8. Added the rcpt_4xx retry error code.
2836 9. Added postmaster_mailfrom=xxx to callout verification option.
2838 10. Added mailfrom=xxxx to the callout verification option, for verify=
2841 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2842 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2844 12. Inserted some casts to stop certain compilers warning when using pointer
2845 differences as field lengths or precisions in printf-type calls (mostly
2846 affecting debugging statements).
2848 13. Added optional readline() support for -be (dynamically loaded).
2850 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2851 same clock tick as a message's arrival, so that its received time was the
2852 same as the "first fail" time on the retry record, and that message
2853 remained on the queue past the ultimate address timeout, every queue runner
2854 would try a delivery (because it was past the ultimate address timeout) but
2855 after another failure, the ultimate address timeout, which should have then
2856 bounced the address, did not kick in. This was a "< instead of <=" error;
2857 in most cases the first failure would have been in the next clock tick
2858 after the received time, and all would be well.
2860 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2861 being recognized when the domain list was tested by the match_domain
2862 condition in an expansion string.
2864 16. Added the ${str2b64: operator.
2866 17. Exim was always calling setrlimit() to set a large limit for the number of
2867 processes, without checking whether the existing limit was already
2868 adequate. (It did check for the limit on file descriptors.) Furthermore,
2869 errors from getrlimit() and setrlimit() were being ignored. Now they are
2870 logged to the main and panic logs, but Exim does carry on, to try to do its
2871 job under whatever limits there are.
2873 18. Imported PCRE 5.0.
2875 19. Trivial typo in log message " temporarily refused connection" (the leading
2878 20. If the log selector return_path_on_delivery was set and an address was
2879 redirected to /dev/null, the delivery process crashed because it assumed
2880 that a return path would always be set for a "successful" delivery. In this
2881 case, the whole delivery is bypassed as an optimization, and therefore no
2884 21. Internal re-arrangement: the function for sending a challenge and reading
2885 a response while authentication was assuming a zero-terminated challenge
2886 string. It's now changed to take a pointer and a length, to allow for
2887 binary data in such strings.
2889 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2891 23. Exim was not respecting finduser_retries when seeking the login of the
2892 uid under which it was called; it was always trying 10 times. (The default
2893 setting of finduser_retries is zero.) Also, it was sleeping after the final
2894 failure, which is pointless.
2896 24. Implemented tls_on_connect_ports.
2898 25. Implemented acl_smtp_predata.
2900 26. If the domain in control=submission is set empty, Exim assumes that the
2901 authenticated id is a complete email address when it generates From: or
2902 Sender: header lines.
2904 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2905 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2906 chown and chgrp in /bin and hostname in /usr/bin.
2908 28. Exim was keeping the "process log" file open after each use, just as it
2909 does for the main log. This opens the possibility of it remaining open for
2910 long periods when the USR1 signal hits a daemon. Occasional processlog
2911 errors were reported, that could have been caused by this. Anyway, it seems
2912 much more sensible not to leave this file open at all, so that is what now
2915 29. The long-running daemon process does not normally write to the log once it
2916 has entered its main loop, and it closes the log before doing so. This is
2917 so that log files can straightforwardly be renamed and moved. However,
2918 there are a couple of unusual error situations where the daemon does write
2919 log entries, and I had neglected to close the log afterwards.
2921 30. The text of an SMTP error response that was received during a remote
2922 delivery was being truncated at 512 bytes. This is too short for some of
2923 the long messages that one sometimes sees. I've increased the limit to
2926 31. It is now possible to make retry rules that apply only when a message has a
2927 specific sender, in particular, an empty sender.
2929 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2930 it possible to be selective about when SMTP synchronization is enforced.
2932 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2934 32. Implemented hosts_connection_nolog.
2936 33. Added an ACL for QUIT.
2938 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2941 35. Added mailbox_size and mailbox_filecount to appendfile.
2943 36. Added control = no_multiline_responses to ACLs.
2945 37. There was a bug in the logic of the code that waits for the clock to tick
2946 in the case where the clock went backwards by a substantial amount such
2947 that the microsecond fraction of "now" was more than the microsecond
2948 fraction of "then" (but the whole seconds number was less).
2950 38. Added support for the libradius Radius client library this is found on
2951 FreeBSD (previously only the radiusclient library was supported).
2957 1. When certain lookups returned multiple values in the form name=value, the
2958 quoting of the values was not always being done properly. Specifically:
2959 (a) If the value started with a double quote, but contained no whitespace,
2961 (b) If the value contained whitespace other than a space character (i.e.
2962 tabs or newlines or carriage returns) it was not quoted.
2963 This fix has been applied to the mysql and pgsql lookups by writing a
2964 separate quoting function and calling it from the lookup code. The fix
2965 should probably also be applied to nisplus, ibase and oracle lookups, but
2966 since I cannot test any of those, I have not disturbed their existing code.
2968 2. A hit in the callout cache for a specific address caused a log line with no
2969 reason for rejecting RCPT. Now it says "Previous (cached) callout
2970 verification failure".
2972 3. There was an off-by-one bug in the queryprogram router. An over-long
2973 return line was truncated at 256 instead of 255 characters, thereby
2974 overflowing its buffer with the terminating zero. As well as fixing this, I
2975 have increased the buffer size to 1024 (and made a note to document this).
2977 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2978 when Exim is waiting for an SMTP response from a remote server, Exim
2979 restarts its select() call on the socket, thereby resetting its timeout.
2980 This is not a problem when such interrupts are rare. Somebody set up a cron
2981 job to run exiwhat every 2 minutes, which is less than the normal select()
2982 timeout (5 or 10 minutes). This meant that the select() timeout never
2983 kicked in because it was always reset. I have fixed this by comparing the
2984 time when an interrupt arrives with the time at the start of the first call
2985 to select(). If more time than the timeout has elapsed, the interrupt is
2986 treated as a timeout.
2988 5. Some internal re-factoring in preparation for the addition of Sieve
2989 extensions (by MH). In particular, the "personal" test is moved to a
2990 separate function, and given an option for scanning Cc: and Bcc: (which is
2991 not set for Exim filters).
2993 6. When Exim created an email address using the login of the caller as the
2994 local part (e.g. when creating a From: or Sender: header line), it was not
2995 quoting the local part when it contained special characters such as @.
2997 7. Installed new OpenBSD configuration files.
2999 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3000 try to make them clearer.
3002 9. Callout options, other than the timeout value, were being ignored when
3003 verifying sender addresses in header lines. For example, when using
3005 verify = header_sender/callout=no_cache
3007 the cache was (incorrectly) being used.
3009 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3010 only the Cygwin environment.
3012 11. When return_path_on_delivery was set as a log selector, if different remote
3013 addresses in the same message used different return paths and parallel
3014 remote delivery occurred, the wrong values would sometimes be logged.
3015 (Whenever a remote delivery process finished, the return path value from
3016 the most recently started remote delivery process was logged.)
3018 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3019 lines when AUTH and/or TLS are in use. This is the "received protocol"
3020 field. Exim used to use "asmtp" for authenticated SMTP, without any
3021 indication (in the protocol name) for TLS use. Now it follows the RFC and
3022 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3023 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3024 names appear in log lines as well as in Received: header lines.
3026 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3027 extensions, and comparison tests, and to fix some bugs.
3029 14. Changes to the "personal" filter test:
3031 (1) The test was buggy in that it was just doing the equivalent of
3032 "contains" tests on header lines. For example, if a user's address was
3033 anne@some.where, the "personal" test would incorrectly be true for
3035 To: susanne@some.where
3037 This test is now done by extracting each address from the header in turn,
3038 and checking the entire address. Other tests that are part of "personal"
3039 are now done using regular expressions (for example, to check local parts
3040 of addresses in From: header lines).
3042 (2) The list of non-personal local parts in From: addresses has been
3043 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3044 taken from the Sieve specification recommendations.
3046 (3) If the message contains any header line starting with "List-" it is
3047 treated as non-personal.
3049 (4) The test for "circular" in the Subject: header line has been removed
3050 because it now seems ill-conceived.
3052 15. Minor typos in src/EDITME comments corrected.
3054 16. Installed latest exipick from John Jetmore.
3056 17. If headers_add on a router specified a text string that was too long for
3057 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3058 of string_sprintf() is now avoided.
3060 18. $message_body_size was not set (it was always zero) when running the DATA
3061 ACL and the local_scan() function.
3063 19. For the "mail" command in an Exim filter, no default was being set for
3064 the once_repeat time, causing a random time value to be used if "once" was
3065 specified. (If the value happened to be <= 0, no repeat happened.) The
3066 default is now 0s, meaning "never repeat". The "vacation" command was OK
3067 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3068 (I found it when inspecting the code).
3070 20. There is now an overall timeout for performing a callout verification. It
3071 defaults to 4 times the callout timeout, which applies to individual SMTP
3072 commands during the callout. The overall timeout applies when there is more
3073 than one host that can be tried. The timeout is checked before trying the
3074 next host. This prevents very long delays if there are a large number of
3075 hosts and all are timing out (e.g. when the network connections are timing
3076 out). The value of the overall timeout can be changed by specifying an
3077 additional sub-option for "callout", called "maxwait". For example:
3079 verify = sender/callout=5s,maxwait=20s
3081 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3082 to the end before writing, but this should make it even safer).
3084 22. Exim was forgetting that it had advertised PIPELINING for the second and
3085 subsequent messages on an SMTP connection. It was also not resetting its
3086 memory on STARTTLS and an internal HELO.
3088 23. When Exim logs an SMTP synchronization error within a session, it now
3089 records whether PIPELINING has been advertised or not.
3091 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3092 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3093 rather than long int.
3095 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3097 26. Added the never_mail option to autoreply.
3103 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3104 crash if the getsockname() call failed; this can happen if a connection is
3105 closed very soon after it is established. The problem was simply in the
3106 order in which certain operations were done, causing Exim to try to write
3107 to the SMTP stream before it had set up the file descriptor. The bug has
3108 been fixed by making things happen in the correct order.
3114 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3115 before the connection was closed, thus losing the rejection response.
3117 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3118 some early Solaris releases, but causes trouble in current releases where
3119 socklen_t is defined.
3121 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3124 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3127 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3128 that releasing the top of it at the end releases what was used for sub-
3129 expansions (unless the block got too big). However, discard this block if
3130 the first thing is a variable or header, so that we can use its block when
3131 it is dynamic (useful for very large $message_headers, for example).
3133 6. Lookups now cache *every* query, not just the most recent. A new, separate
3134 store pool is used for this. It can be recovered when all lookup caches are
3135 flushed. Lookups now release memory at the end of their result strings.
3136 This has involved some general refactoring of the lookup sources.
3138 7. Some code has been added to the store_xxx() functions to reduce the amount
3139 of flapping under certain conditions.
3141 8. log_incoming_interface used to affect only the <= reception log lines. Now
3142 it causes the local interface and port to be added to several more SMTP log
3143 lines, for example "SMTP connection from", and rejection lines.
3145 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3147 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3149 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3150 could be overwritten at the end of the current message (or the start of a
3151 new message if it was set in a HELO ACL). The value is now preserved for
3152 the duration of the SMTP connection.
3154 12. If a transport had a headers_rewrite setting, and a matching header line
3155 contained an unqualified address, that address was qualified, even if it
3156 did not match any rewriting rules. The underlying bug was that the values
3157 of the flags that permit the existence of unqualified sender and recipient
3158 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3159 non-local messages, and by -bnq for local messages) were not being
3160 preserved with the message after it was received.
3162 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3163 "next input=" as part of the text comprising the host identity instead of
3164 the correct text. The code was using the same buffer for two different
3165 strings. However, depending on which order the printing function evaluated
3166 its arguments, the bug did not always show up. Under Linux, for example, my
3167 test suite worked just fine.
3169 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3170 doesn't work with some older versions of Perl. It has been changed to "my",
3171 which in any case is probably the better facility to use.
3173 15. A really picky compiler found some instances of statements for creating
3174 error messages that either had too many or two few arguments for the format
3177 16. The size of the buffer for calls to the DNS resolver has been increased
3178 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3179 for addresses that have a lot of PTR records. This alleviates a problem; it
3180 does not fully solve it.
3182 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3183 buffer now truncates the list and logs the incident, which is the same
3184 action as happens when Exim is looking up a host name and its aliases.
3185 Previously in this situation something unpredictable would happen;
3186 sometimes it was "internal error: store_reset failed".
3188 18. If a server dropped the connection unexpectedly when an Exim client was
3189 using GnuTLS and trying to read a response, the client delivery process
3190 crashed while trying to generate an error log message.
3192 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3193 string, for example:
3195 warn message = H1: something\nH2: something
3197 the text was added as a single header line from Exim's point of view
3198 though it ended up OK in the delivered message. However, searching for the
3199 second and subsequent header lines using $h_h2: did not work. This has been
3200 fixed. Similarly, if a system filter added multiple headers in this way,
3201 the routers could not see them.
3203 20. Expanded the error message when iplsearch is called with an invalid key to
3204 suggest using net-iplsearch in a host list.
3206 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3207 ACLs are no longer actually imposed (and a message to that effect is
3210 22. If a "gecos" field in a passwd entry contained escaped characters, in
3211 particular, if it contained a \" sequence, Exim got it wrong when building
3212 a From: or a Sender: header from that name. A second bug also caused
3213 incorrect handling when an unquoted " was present following a character
3214 that needed quoting.
3216 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3217 was not being matched caselessly.
3219 24. Arranged for all hyphens in the exim.8 source to be escaped with
3222 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3223 properly. Recipient callout cache records were still being keyed to include
3224 the sender, even when use_sender was set false. This led to far more
3225 callouts that were necessary. The sender is no longer included in the key
3226 when use_sender is false.
3228 26. Added "control = submission" modifier to ACLs.
3230 27. Added the ${base62d: operator to decode base 62 numbers.
3232 28. dnsdb lookups can now access SRV records.
3234 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3235 the configuration file.
3237 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3238 (-v) mode. This makes the output for a verbose queue run more intelligible.
3240 31. Added a use_postmaster feature to recipient callouts.
3242 32. Added the $body_zerocount variable, containing the number of binary zero
3243 bytes in the message body.
3245 33. The time of last modification of the "new" subdirectory is now used as the
3246 "mailbox time last read" when there is a quota error for a maildir
3249 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3251 35. Added +ignore_unknown as a special item in host lists.
3253 36. Code for decoding IPv6 addresses in host lists is now included, even if
3254 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3255 address was recognized as an IP address, but was then not correctly decoded
3256 into binary, causing unexpected and incorrect effects when compared with
3263 1. Very minor rewording of debugging text in manualroute to say "list of
3264 hosts" instead of "hostlist".
3266 2. If verify=header_syntax was set, and a header line with an unqualified
3267 address (no domain) and a large number of spaces between the end of the
3268 name and the colon was received, the reception process suffered a buffer
3269 overflow, and (when I tested it) crashed. This was caused by some obsolete
3270 code that should have been removed. The fix is to remove it!
3272 3. When running in the test harness, delay a bit after writing a bounce
3273 message to get a bit more predictability in the log output.
3275 4. Added a call to search_tidyup() just before forking a reception process. In
3276 theory, someone could use a lookup in the expansion of smtp_accept_max_
3277 per_host which, without the tidyup, could leave open a database connection.
3279 5. Added the variables $recipient_data and $sender_data which get set from a
3280 lookup success in an ACL "recipients" or "senders" condition, or a router
3281 "senders" option, similar to $domain_data and $local_part_data.
3283 6. Moved the writing of debug_print from before to after the "senders" test
3286 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3287 problems for message scanning, either using a data ACL, or using
3288 local_scan() because the Received: header was not generated till after they
3289 were called (in order to set the time as the time of reception completion).
3290 I have revised the way this works. The header is now generated after the
3291 body is received, but before the ACL or local_scan() are called. After they
3292 are run, the timestamp in the header is updated.
3298 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3299 before starting a queue runner without re-exec. This happened only when
3300 deliver_drop_privilege was set or when the Exim user was set to root. The
3301 effect of the bug was that timeouts during subsequent deliveries caused
3302 crashes instead of being properly handled. The handler is now left at its
3303 default (and expected) setting.
3305 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3306 message, again when deliver_drop_privilege is set or Exim is run as root.
3307 The bug described in (1) was not present in this case, but the tidying up
3308 of the other signals was missing. I have made the two cases consistent.
3310 3. The ignore_target_hosts setting on a manualroute router was being ignored
3311 for hosts that were looked up using the /MX notation.
3313 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3316 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3317 operated on the sender address. After changing the $sender_address to <>
3318 for the sender address verify, Exim was re-instated it as the original
3319 (before rewriting) address, but remembering that it had rewritten it, so it
3320 wasn't rewriting it again. This bug also had the effect of breaking the
3321 sender address verification caching when the sender address was rewritten.
3323 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3324 This has been changed so that if the ip literal address matches
3325 ignore_target_hosts, the router declines.
3327 7. Added expansion conditions match_domain, match_address, and match_local_
3328 part (NOT match_host).
3330 8. The placeholder for the Received: header didn't have a length field set.
3332 9. Added code to Exim itself and to exim_lock to test for a specific race
3333 condition that could lead to file corruption when using MBX delivery. The
3334 issue is with the lockfile that is created in /tmp. If this file is removed
3335 after a process has opened it but before that process has acquired a lock,
3336 there is the potential for a second process to recreate the file and also
3337 acquire a lock. This could lead to two Exim processes writing to the file
3338 at the same time. The added code performs the same test as UW imapd; it
3339 checks after acquiring the lock that its file descriptor still refers to
3340 the same named file.
3342 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3343 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3346 11. Added the smtp_active_hostname option. If used, this will typically be made
3347 to depend on the incoming interface address. Because $interface_address is
3348 not set up until the daemon has forked a reception process, error responses
3349 that can happen earlier (such as "too many connections") no longer contain
3352 12. If an expansion in a condition on a "warn" statement fails because a lookup
3353 defers, the "warn" statement is abandoned, and the next ACL statement is
3354 processed. Previously this caused the whole ACL to be aborted.
3356 13. Added the iplsearch lookup type.
3358 14. Added ident_timeout as a log selector.
3360 15. Added tls_certificate_verified as a log selector.
3362 16. Added a global option tls_require_ciphers (compare the smtp transport
3363 option of the same name). This controls incoming TLS connections.
3365 17. I finally figured out how to make tls_require_ciphers do a similar thing
3366 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3367 before starting the TLS session.
3369 18. Tabs are now shown as \t in -bP output.
3371 19. If the log selector return_path_on_delivery was set, Exim crashed when
3372 bouncing a message because it had too many Received: header lines.
3374 20. If two routers both had headers_remove settings, and the first one included
3375 a superfluous trailing colon, the final name in the first list and the
3376 first name in the second list were incorrectly joined into one item (with a
3377 colon in the middle).
3383 1. Added -C and -D options to the exinext utility, mainly to make it easier
3384 to include in the automated testing, but these could be helpful when
3385 multiple configurations are in use.
3387 2. The exinext utility was not formatting the output nicely when there was
3388 an alternate port involved in the retry record key, nor when there was a
3389 message id as well (for retries that were specific to a specific message
3390 and a specific host). It was also confused by IPv6 addresses, because of
3391 the additional colons they contain. I have fixed the IPv4 problem, and
3392 patched it up to do a reasonable job for IPv6.
3394 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3395 delivery, the log line now contains "pipelined" if PIPELINING was used.
3397 4. An SMTP transport process used to panic and die if the bind() call to set
3398 an explicit outgoing interface failed. This has been changed; it is now
3399 treated in the same way as a connect() failure.
3401 5. A reference to $sender_host_name in the part of a conditional expansion
3402 that was being skipped was still causing a DNS lookup. This no longer
3405 6. The def: expansion condition was not recognizing references to header lines
3406 that used bh_ and bheader_.
3408 7. Added the _cache feature to named lists.
3410 8. The code for checking quota_filecount in the appendfile transport was
3411 allowing one more file than it should have been.
3413 9. For compatibility with Sendmail, the command line option
3421 and sets the incoming protocol and host name (for trusted callers). The
3422 host name and its colon can be omitted when only the protocol is to be set.
3423 Note the Exim already has two private options, -pd and -ps, that refer to
3424 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3425 "s", but I don't think that's a major issue.
3427 10. A number of refactoring changes to the code, none of which should affect
3430 (a) The number of logging options was getting close to filling up the
3431 32-bit word that was used as a bit map. I have split them into two classes:
3432 those that are passed in the argument to log_write(), and those that are
3433 only ever tested independently outside of that function. These are now in
3434 separate 32-bit words, so there is plenty of room for expansion again.
3435 There is no change in the user interface or the logging behaviour.
3437 (b) When building, for example, log lines, the code previously used a
3438 macro that called string_cat() twice, in order to add two strings. This is
3439 not really sufficiently general. Furthermore, there was one instance where
3440 it was actually wrong because one of the argument was used twice, and in
3441 one call a function was used. (As it happened, calling the function twice
3442 did not affect the overall behaviour.) The macro has been replaced by a
3443 function that can join an arbitrary number of extra strings onto a growing
3446 (c) The code for expansion conditions now uses a table and a binary chop
3447 instead of a serial search (which was left over from when there were very
3448 few conditions). Also, it now recognizes conditions like "pam" even when
3449 the relevant support is not compiled in: a suitably worded error message is
3450 given if an attempt is made to use such a condition.
3452 11. Added ${time_interval:xxxxx}.
3454 12. A bug was causing one of the ddress fields not to be passed back correctly
3455 from remote delivery subprocesses. The field in question was not being
3456 subsequently used, so this caused to problems in practice.
3458 13. Added new log selectors queue_time and deliver_time.
3460 14. Might have fixed a bug in maildirsizefile handling that threw up
3461 "unexpected character" debug warnings, and recalculated the data
3462 unnecessarily. In any case, I expanded the warning message to give more
3465 15. Added the message "Restricted characters in address" to the statements in
3466 the default ACL that block characters like @ and % in local parts.
3468 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3469 Three changes have been made:
3471 (a) There was a serious bug; a negative response to MAIL caused the whole
3472 recipient domain to be cached as invalid, thereby blocking all messages
3473 to all local parts at the same domain, from all senders. This bug has
3474 been fixed. The domain is no longer cached after a negative response to
3475 MAIL if the sender used is not empty.
3477 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3480 (c) A new callout option, "use_sender" has been added for people who want
3481 the modified behaviour.
3487 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3490 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3491 indeed breaks things for older releases.
3493 3. Added additional logging to the case where there is a problem reading data
3494 from a filter that is running in a subprocess using a pipe, in order to
3495 try to track down a specific problem.
3497 4. Testing facility fudge: when running in the test harness and attempting
3498 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3499 getting "No route to host". Convert this to a timeout.
3501 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3504 6. Some OS don't have socklen_t but use size_t instead. This affects the
3505 fifth argument of getsockopt() amongst other things. This is now
3506 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3507 can be set for individual OS. I have set it for SunOS5, OSF1, and
3508 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3509 some earlier ones do not.
3511 7. Change 4.30/15 was not doing the test caselessly.
3513 8. The standard form for an IPv6 address literal was being rejected by address
3514 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3515 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3516 this, as well as the form without the "IPv6" on the front (but only when
3517 address literals are enabled, of course).
3519 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3521 10. Exim crashed if a message with an empty sender address specified by -f
3522 encountered a router with an errors_to setting. This could be provoked only
3523 by a command such as
3527 where an empty string was supplied; "<>" did not hit this bug.
3529 11. Installed PCRE release 4.5.
3531 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3532 remained set. It is now erased.
3534 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3535 times from message ids (which are base 36 rather than the normal 62).
3537 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3538 were being counted as actual protocol errors, and logged if the log
3539 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3540 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3541 and DATA following a set of rejected RCPTs do not count as protocol errors.
3542 In other words, Exim assumes they were pipelined, though this may not
3543 actually be the case. Of course, in all cases the client gets an
3544 appropriate error code.
3546 15. If a lookup fails in an ACL condition, a message about the failure may
3547 be available; it is used if testing the ACL cannot continue, because most
3548 such messages specify what the cause of the deferral is. However, some
3549 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3550 that caused an old message to be retained and used if a later statement
3551 caused a defer, replacing the real cause of the deferral.
3553 16. If an IP address had so many PTR records that the DNS lookup buffer
3554 was not large enough to hold them, Exim could crash while trying to process
3555 the truncated data. It now detects and logs this case.
3557 17. Further to 4.21/58, another change has been made: if (and only if) the
3558 first line of a message (the first header line) ends with CRLF, a bare LF
3559 in a subsequent header line has a space inserted after it, so as not to
3560 terminate the header.
3562 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3563 unnecessarily, used atoi() instead of strtol(), and didn't check the
3564 termination when getting file sizes from file names by regex.
3566 19. Completely re-implemented the support for maildirsize files, in the light
3567 of a number of problems with the previous contributed implementation
3568 (4.30/29). In particular:
3570 . If the quota is zero, the maildirsize file is maintained, but no quota is
3573 . If the maildir directory does not exist, it is created before any attempt
3574 to write a maildirsize file.
3576 . The quota value in the file is just a cache; if the quota is changed in
3577 the transport, the new value overrides.
3579 . A regular expression is available for excluding directories from the
3582 20. The autoreply transport checks the characters in options that define the
3583 message's headers; it allows continued headers, but it was checking with
3584 isspace() after an embedded newline instead of explicitly looking for a
3587 21. If all the "regular" hosts to which an address was routed had passed their
3588 expiry times, and had not reached their retry times, the address was
3589 bounced, even if fallback hosts were defined. Now Exim should go on to try
3592 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3593 equivalent code in the SMTP transport. Some hosts send humungous responses
3594 to HELO/EHLO, more than 1024 it seems.
3596 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3597 gives compiler warnings in some environments. I've now done it "properly",
3600 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3601 (because of problems with the built-in one) was declared to return uschar *
3602 instead of char *, causing compiler failure.
3604 25. Fixed a file descriptor leak when processing alias/forward files.
3606 26. Fixed a minor format string issue in dbfn.c.
3608 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3610 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3611 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3614 29. When a sender address is verified, it is cached, to save repeating the test
3615 when there is more than one recipient in a message. However, when the
3616 verification involves a callout, it is possible for different callout
3617 options to be set for different recipients. It is too complicated to keep
3618 track of this in the cache, so now Exim always runs a verification when a
3619 callout is required, relying on the callout cache for the optimization.
3620 The overhead is duplication of the address routing, but this should not be
3623 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3624 to be verified with callout=postmaster, and the main callout worked but the
3625 postmaster check failed, the verification correctly failed. However, if a
3626 subsequent RCPT command asked for sender verification *without* the
3627 postmaster check, incorrect caching caused this verification also to fail,
3630 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3631 it was not caching the DNS options (qualify_single, search_parents) that
3632 were used when the lookup failed. A subsequent lookup with different
3633 options therefore always gave the same answer, though there were cases
3634 where it should not have. (Example: a "domains = !$mx_any" option on a
3635 dnslookup router: the "domains" option is always processed without any
3636 widening, but the router might have qualify_single set.) Now Exim uses the
3637 cached value only when the same options are set.
3639 32. Added John Jetmore's "exipick" utility to the distribution.
3641 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3642 than a timeout (e.g. a certificate is required, and is not provided), an
3643 Exim server now closes the connection immediately. Previously it waited for
3644 the client to close - but if the client is SSL, it seems that they each
3645 wait for each other, leading to a delay before one of them times out.
3647 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3648 maintained 0.8.x compatibility because I don't think many are using it, and
3649 it is clearly obsolete.
3651 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3654 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3655 client certificate was expired. A simple patch fixes this, though I don't
3656 understand the full logic of why the verify callback is called multiple
3659 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3660 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3661 which causes problems with some clients (such as the Certicom SSL Plus
3662 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3663 disables the coutermeasure allowing Eudora to connect."
3665 38. Exim was not checking that a write() to a log file succeeded. This could
3666 lead to Bad Things if a log got too big, in particular if it hit a file
3667 size limit. Exim now panics and dies if it cannot write to a log file, just
3668 as it does if it cannot open a log file.
3670 39. Modified OS/Makefile-Linux so that it now contains
3672 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3674 The two -D definitions ensure that Exim is compiled with large file
3675 support, which makes it possible to handle log files that are bigger than
3678 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3679 instance) a domain was checked against a named list that involved a lookup,
3680 causing $domain_data to be set, then another domain was checked against the
3681 same list, then the first domain was re-checked, the value of $domain_data
3682 after the final check could be wrong. In particular, if the second check
3683 failed, it could be set empty. This bug probably also applied to
3686 41. The strip_trailing_dot option was not being applied to the address given
3687 with the -f command-line option.
3689 42. The code for reading a message's header from the spool was incrementing
3690 $received_count, but never initializing it. This meant that the value was
3691 incorrect (doubled) while delivering a message in the same process in which
3692 it was received. In the most common configuration of Exim, this never
3693 happens - a fresh exec is done - but it can happen when
3694 deliver_drop_privilege is set.
3696 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3697 it now includes up to 150 characters of the unexpected data in the log
3700 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3701 and building data strings. The size of both of these buffers was 10 000
3702 bytes - far larger than anybody would *ever* want, thought I. Needless to
3703 say, somebody hit the limit. I have increased the maximum line length to
3704 20 000 and the maximum data length of concatenated lines to 100 000. I have
3705 also fixed two bugs, because there was no checking on these buffers. Tsk,
3706 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3707 buffer is too small.
3709 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3710 lsearch lookups. Now it does.
3712 46. When parsing a route_list item in a manualroute router, a fixed-length
3713 buffer was used for the list of hosts. I made this 1024 bytes long,
3714 thinking that nobody would ever have a list of hosts that long. Wrong.
3715 Somebody had a whole pile of complicated expansion conditions, and the
3716 string was silently truncated, leading to an expansion error. It turns out
3717 that it is easier to change to an unlimited length (owing to other changes
3718 that have happened since this code was originally written) than to build
3719 structure for giving a limitation error. The length of the item that
3720 expands into the list of hosts is now unlimited.
3722 47. The lsearch lookup could not handle data where the length of text line was
3723 more than 4095 characters. Such lines were truncated, leading to shortened
3724 data being returned. It should now handle lines of any length.
3726 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3727 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3730 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3731 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3732 no longer happens in comment lines. A list of replacements is now placed
3733 at the head of all of the source files, except those whose only change is
3734 to replace PERL_COMMAND in the very first #! line.
3736 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3737 messages on the queue, with a bottom-up merge sort, using code contributed
3738 by Michael Haardt. This should make operations like -bp somewhat faster on
3739 large queues. It won't affect queue runners, except when queue_run_in_order
3742 51. Installed eximstats 1.31 in the distribution.
3744 52. Added support for SRV lookups to the dnslookup router.
3746 53. If an ACL referred to $message_body or $message_body_end, the value was not
3747 reset for any messages that followed in the same SMTP session.
3749 54. The store-handling optimization for building very long strings was not
3750 differentiating between the different store pools. I don't think this
3751 actually made any difference in practice, but I've tidied it.
3753 55. While running the routers to verify a sender address, $sender_address
3754 was still set to the sender address. This is wrong, because when routing to
3755 send a bounce to the sender, it would be empty. Therefore, I have changed
3756 it so that, while verifying a sender address, $sender_address is set to <>.
3757 (There is no change to what happens when verifying a recipient address.)
3759 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3760 target A or AAAA records (if not already returned) without resetting the
3761 qualify_single or search_parents options of the DNS resolver. These are
3762 inappropriate in this case because the targets of MX and SRV records must
3763 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3764 target that, when qualified, matched something in the local domain. These
3765 two options are now turned off when doing these lookups.
3767 57. It seems that at least some releases of Reiserfs (which does not have the
3768 concept of a fixed number of inodes) returns zero and not -1 for the
3769 number of available inodes. This interacted badly with check_spool_inodes,
3770 which assumed that -1 was the "no such thing" setting. What I have done is
3771 to check that the total number of inodes is greater than zero before doing
3772 the test of how many are available.
3774 58. When a "warn" ACL statement has a log_message modifier, the message is
3775 remembered, and not repeated. This is to avoid a lot of repetition when a
3776 message has many recipients that cause the same warning to be written.
3777 Howewer, Exim was preserving the list of already written lines for an
3778 entire SMTP session, which doesn't seem right. The memory is now reset if a
3779 new message is started.
3781 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3782 debugging output unless log_rewrite was also set.
3784 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3785 of (int)(handle) when we know that handle contains (void *)(-1).
3787 61. The Exim daemon panic-logs an error return when it closes the incoming
3788 connection. However "connection reset by peer" seems to be common, and
3789 isn't really an error worthy of noting specially, so that particular error
3792 62. When Exim is trying to find all the local interfaces, it used to panic and
3793 die if the ioctl to get the interface flags failed. However, it seems that
3794 on at least one OS (Solaris 9) it is possible to have an interface that is
3795 included in the list of interfaces, but for which you get a failure error
3796 for this call. This happens when the interface is not "plumbed" into a
3797 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3798 failure of the "get flags" call assumes that the interface is down.
3800 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3801 makes life easier for people who are doing arithmetic on fields extracted
3802 from dates, where you often get leading zeros that should not be
3803 interpreted as octal.
3805 64. Added qualify_domain to the redirect router, to override the global
3808 65. If a pathologically long header line contained very many addresses (the
3809 report of this problem mentioned 10 000) and each of them was rewritten,
3810 Exim could use up a very large amount of memory. (It kept on making new
3811 copies of the header line as it rewrote, and never released the old ones.)
3812 At the expense of a bit more processing, the header rewriting function has
3813 been changed so that it no longer eats memory in this way.
3815 66. The generation of the Received: header has been moved from the time that a
3816 message starts to be received, to the time that it finishes. The timestamp
3817 in the Received: header should now be very close to that of the <= log
3818 line. There are two side-effects of this change:
3820 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3821 logged header lines no longer include the local Received: line, because
3822 it has not yet been created. The same applies to a copy of the message
3823 that is returned to a non-SMTP sender when a message is rejected.
3825 (b) When a filter file is tested using -bf, no additional Received: header
3826 is added to the test message. After some thought, I decided that this
3829 This change does not affect the value of $received_for. It is still set
3830 after address rewriting, but before local_scan() is called.
3832 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3834 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3835 gave an unhelpful panic error message, and a defer error. I have managed to
3836 change this behaviour so that it now rejects any supplied certificate,
3837 which seems right, as the list of acceptable certificates is empty.
3839 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3840 gave an unhelpful defer error. I have not managed to make this reject any
3841 supplied certificates, but the error message it gives is "no certificate
3842 supplied", which is not helpful.
3844 70. exigrep's output now also includes lines that are not associated with any
3845 message, but which match the given pattern. Implemented by a patch from
3846 Martin Sluka, which also tidied up the Perl a bit.
3848 71. Recipient callout verification, like sender verification, was using <> in
3849 the MAIL FROM command. This isn't really the right thing, since the actual
3850 sender may affect whether the remote host accepts the recipient or not. I
3851 have changed it to use the actual sender in the callout; this means that
3852 the cache record is now keyed on a recipient/sender pair, not just the
3853 recipient address. There doesn't seem to be a real danger of callout loops,
3854 since a callout by the remote host to check the sender would use <>.
3855 [SEE ABOVE: changed after hitting problems.]
3857 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3858 temporary errors. However, in the case of such a code being given after
3859 the end of a data transmission (i.e. after ".") Exim was failing to write
3860 a retry record for the message. (Yes, there was some broken host that was
3861 actually sending 8xx at this point.)
3863 73. An unknown lookup type in a host list could cause Exim to panic-die when
3864 the list was checked. (An example that provoked this was putting <; in the
3865 middle of a list instead of at the start.) If this happened during a DATA
3866 ACL check, a -D file could be left lying around. This kind of configuration
3867 error no longer causes Exim to die; instead it causes a defer errror. The
3868 incident is still logged to the main and panic logs.
3870 74. Buglet left over from Exim 3 conversion. The message "too many messages
3871 in one connection" was written to the rejectlog but not the mainlog, except
3872 when address rewriting (yes!) was being logged.
3874 75. Added write_rejectlog option.
3876 76. When a system filter was run not as root (that is, when system_filter_user
3877 was set), the values of the $n variables were not being returned to the
3878 main process; thus, they were not subsequently available in the $sn
3881 77. Added +return_path_on_delivery log selector.
3883 78. A connection timeout was being treated differently from recipients deferred
3884 when testing hosts_max_try with a message that was older than the host's
3885 retry timeout. (The host should not be counted, thus allowing all hosts to
3886 be tried at least once before bouncing.) This may have been the cause of an
3887 occasionally reported bug whereby a message would remain on the queue
3888 longer than the retry timeout, but would be bounced if a delivery was
3889 forced. I say "may" because I never totally pinned down the problem;
3890 setting up timeout/retry tests is difficult. See also the next item.
3892 79. The ultimate address timeout was not being applied to errors that involved
3893 a combination of host plus message (for example, a timeout on a MAIL
3894 command). When an address resolved to a number of possible hosts, and they
3895 were not all tried for each delivery (e.g. because of hosts_max_try), a
3896 message could remain on the queue longer than the retry timeout.
3898 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3901 81. Fixed an obscure SMTP outgoing bug which required at least the following
3902 conditions: (a) there was another message waiting for the same server;
3903 (b) the server returned 5xx to all RCPT commands in the first message so
3904 that the message was not completed; (c) the server dropped the connection
3905 or gave a negative response to the RSET that Exim sends to abort the
3906 transaction. The observed case was a dropped connection after DATA that had
3907 been sent in pipelining mode. That is, the server had advertised PIPELINING
3908 but was not implementing it correctly. The effect of the bug was incorrect
3909 behaviour, such as trying another host, and this could lead to a crash.
3915 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3916 and daemon.c were passed as pointers to ints; they should have been
3917 pointers to socklen_t variables (which are typically unsigned ints).
3919 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3922 3. Fixed a really odd bug that affected only the testing scheme; patching a
3923 certain fixed string in the binary changed the value of another string that
3924 happened to be identical to the end of the original first string.
3926 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3927 name", it returns that address as the IP address. On some operating
3928 systems (e.g. Solaris), it also passes back the IP address string as the
3929 "host name". However, on others (e.g. Linux), it passes back an empty
3930 string. Exim wasn't checking for this, and was changing the host name to an
3931 empty string, assuming it had been canonicized.
3933 5. Although rare, it is permitted to have more than one PTR record for a given
3934 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3935 all the names associated with an address, because they do in Solaris.
3936 However, it seems that they do not in Linux for data that comes from the
3937 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3938 I found this out when I moved to a new Linux workstation and tried to run
3939 the Exim test suite.
3941 To get round this problem I have changed the code so that it now does its
3942 own call to the DNS to look up PTR records when searching for a host name.
3943 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3944 addresses that are only in /etc/hosts are still found.
3946 This behaviour is, however, controlled by an option called host_lookup_
3947 order, which defaults to "bydns:byaddr". If people want to use the other
3948 order, or indeed, just use one or the other means of lookup, they can
3949 specify it in this variable.
3951 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3952 some operating systems, this comes back from gethostbyaddr() as an empty
3953 string, and this is what Exim used to test for. However, it seems that in
3954 other systems, "." is yielded. Exim now tests for this case too.
3956 7. The values of check_spool_space and check_log_space are now held internally
3957 as a number of kilobytes instead of an absolute number of bytes. If a
3958 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3959 kilobyte. This means that much larger values can be stored.
3961 8. Exim monitor: an attempt to get the action menu when not actually pointing
3962 at a message produces an empty menu entitled "No message selected". This
3963 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3964 no entries in it ("Shell widget menu has zero width and/or height"). So I
3965 have added a single, blank menu entry in this case.
3967 9. Added ${quote_local_part.
3969 10. MIME decoding is now applied to the contents of Subject: header lines when
3972 11. Now that a reference to $sender_host_address automatically causes a reverse
3973 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3974 host lookup before query-style lookups in lists that might use this
3975 variable. This has therefore been abolished, and the "net-" prefix is no
3976 longer necessary for query-style lookups.
3978 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3979 have been a typo for LFLAGS, so it has been changed.
3981 13. The install script calls Exim with "-C /dev/null" in order to find the
3982 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3983 to be output. Howeve, since Exim outputs its version number before the
3984 error, it didn't break the script. It just looked ugly. I fixed this by
3985 always allowing "-C /dev/null" if the caller is root.
3987 14. Ignore overlarge ACL variable number when reading spool file - insurance
3988 against a later release with more variables having written the file.
3990 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3991 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3992 this, as well as the form without the "IPv6" on the front.
3994 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3995 OS/Makefile-Darwin file.
3997 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3998 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4000 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4001 variable. However, some OS use ints for inodes. I've added cast to long int
4002 to get rid of the compiler warning.
4004 19. I had forgotten to lock out "/../" in configuration file names when
4005 ALT_CONFIG_PREFIX was set.
4007 20. Routers used for verification do not need to specify transports. However,
4008 if such a router generated a host list, and callout was configured, Exim
4009 crashed, because it could not find a port number from the (non-existent)
4010 transport. It now assumes port 25 in this circumstance.
4012 21. Added the -t option to exigrep.
4014 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4015 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4016 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4017 them is set, LOOKUP_LSEARCH is forced.
4019 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4021 24. Added sender and host information to the "rejected by local_scan()" log
4022 line; previously there was no indication of these.
4024 25. Added .include_if_exists.
4026 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4027 Linux. It turns out that not all file systems support this. Apparently some
4028 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4029 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4030 not supported on the file descriptor, is now ignored when Exim is trying to
4031 sync a directory. This applies only to Linux.
4033 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4035 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4036 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4039 29. Added support for maildirsize files from supplied patch (modified a bit).
4041 30. The use of :fail: followed by an empty string could lead Exim to respond to
4042 sender verification failures with (e.g.):
4044 550 Verification failed for <xxx>
4045 550 Sender verify failed
4047 where the first response line was missing the '-' that indicates it is not
4048 the final line of the response.
4050 31. The loop for finding the name of the user that called Exim had a hardwired
4051 limit of 10; it now uses the value of finduser_retries, which is used for
4052 all other user lookups.
4054 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4057 33. Exim was neglecting to zero errno before one call of strtol() when
4058 expanding a string and expecting an integer value. On some systems this
4059 resulted in spurious "integer overflow" errors. Also, it was casting the
4060 result into an int without checking.
4062 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4063 did not work. The code looks as if it has *never* worked, though it appears
4064 to have been documented since at least releast 1.62. I have made it work.
4066 35. The "timeout_DNS" error in retry rules, also documented since at least
4067 1.62, also never worked. As it isn't clear exactly what this means, and
4068 clearly it isn't a major issue, I have abolished the feature by treating it
4069 as "timeout", and writing a warning to the main and panic logs.
4071 36. The display of retry rules for -brt wasn't always showing the error code
4074 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4075 timeout_connect_A, timeout_connect_MX.
4077 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4078 to the empty sender.
4080 39. The daemon was not analysing the content of -oX till after it had closed
4081 stderr and disconnected from the controlling terminal. This meant that any
4082 syntax errors were only noted on the panic log, and the return code from
4083 the command was 0. By re-arranging the code a little, I've made the
4084 decoding happen first, so such errors now appear on stderr, and the return
4085 code is 1. However, the actual setting up of the sockets still happens in
4086 the disconnected process, so errors there are still only recorded on the
4089 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4090 connections (as happens on some IP stacks) was logged at start up time as
4091 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4092 it from "IPv6 and IPv4", which means that two separate sockets are being
4095 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4096 says whether AF_INET or AF_INET6 was passed as an argument.
4098 42. Exiwhat output was messed up when time zones were included in log
4101 43. Exiwhat now gives more information about the daemon's listening ports,
4102 and whether -tls-on-connect was used.
4104 44. The "port" option of the smtp transport is now expanded.
4106 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4107 silently ignored. Now an error message is written to the main and panic
4110 46. There's a new ACL modifier called "logwrite" which writes to a log file
4111 as soon as it is encountered.
4113 47. Added $local_user_uid and $local_user_gid at routing time.
4115 48. Exim crashed when trying to verify a sender address that was being
4118 49. Exim was recognizing only a space character after ".include". It now also
4119 recognizes a tab character.
4121 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4122 extracting the relevant information from the specification. The man page no
4123 longer contains scrambled data for the -d option, and I've added a section
4124 at the front about calling Exim under different names.
4126 51. Added "extra_headers" argument to the "mail" command in filter files.
4128 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4131 53. Installed eximstats 1.29.
4133 54. Added transport_filter_timeout as a generic transport option.
4135 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4136 Cc: header lines. This was required by RFC 822, but it not required by RFC
4139 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4140 incoming messages that did not have them. Now it does so only if the
4141 message originates locally, that is, if there is no associated remote host
4142 address. When Resent- header lines are present, this applies to the Resent-
4143 lines rather than the non-Resent- lines.
4145 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4146 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4148 58. Messages for configuration errors now include the name of the main
4149 configuration file - useful now that there may be more than one file in a
4150 list (.included file names were always shown).
4152 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4153 for those rare installations that do not start the daemon as root or run it
4154 setuid root. I've cut out the call to initgroups() if the daemon is not
4157 60. The Exim user and group can now be bound into the binary as text strings
4158 that are looked up at the start of Exim's processing.
4160 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4162 62. Added $mailstore_basename variable.
4164 63. Installed patch to sieve.c from Michael Haardt.
4166 64. When Exim failed to open the panic log after failing to open the main log,
4167 the original message it was trying to log was written to stderr and debug
4168 output, but if they were not available (the usual case in production), it
4169 was lost. Now it is written to syslog before the two lines that record the
4170 failures to open the logs.
4172 65. Users' Exim filters run in subprocesses under the user's uid. It is
4173 possible for a "deliver" command or an alias in a "personal" command to
4174 provoke an address rewrite. If logging of address rewriting is configured,
4175 this fails because the process is not running as root or exim. There may be
4176 a better way of dealing with this, but for the moment (because 4.30 needs
4177 to be released), I have disabled address rewrite logging when running a
4178 filter in a non-root, non-exim process.
4184 1. The buildconfig auxiliary program wasn't quoting the value set for
4185 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4186 not defined. This bug was masked in 4.22 by the effect that was fixed in
4189 2. Some messages that were rejected after a message id was allocated were
4190 shown as "incomplete" by exigrep. It no longer does this for messages that
4191 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4193 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4194 have allow_domain_literals set, the ID did not get logged in the <= line.
4195 Domain literals are now always recognized in Message-ID: header lines.
4197 4. The first argument for a ${extract expansion item is the key name or field
4198 number. Leading and trailing spaces in this item were not being ignored,
4199 causing some misleading effects.
4201 5. When deliver_drop_privilege was set, single queue runner processes started
4202 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4203 same command in the process it spins off) were not dropping privilege.
4205 6. When the daemon running as "exim" started a queue runner, it always
4206 re-executed Exim in the spun-off process. This is a waste of effort when
4207 deliver_drop_privilege is set. The new process now just calls the
4208 queue-runner function directly.
4214 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4217 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4218 ignored. Though the use of -oP was forcing the writing of a pid file, it
4219 was always written to the default place.
4221 3. If the message "no IP address found for host xxxx" is generated during
4222 incoming verification, it is now followed by identification of the incoming
4223 connection (so you can more easily find what provoked it).
4225 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4227 5. Added some features to "harden" Exim a bit more against certain attacks:
4229 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4230 be put in Local/Makefile. This is like the never_users runtime option,
4231 but it cannot be overridden. The default setting is "root".
4233 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4234 prefix string with which any file named in a -C command line option
4237 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4238 is retained for -C and -D only if the caller of Exim is root. Without
4239 it, the exim user may also use -C and -D and retain privilege.
4241 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4242 command line option is disabled.
4244 6. Macro names set by the -D option must start with an upper case letter, just
4245 like macro names defined in the configuration file.
4247 7. Added "dereference=" facility to LDAP.
4249 8. Two instances of the typo "uknown" in the source files are fixed.
4251 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4252 the Configure-Makefile script screwed up while processing it.
4254 10. Incorporated PCRE 4.4.
4256 11. The SMTP synchronization check was not operating right at the start of an
4257 SMTP session. For example, it could not catch a HELO sent before the client
4258 waited for the greeting. There is now a check for outstanding input at the
4259 point when the greeting is written. Because of the duplex, asynchronous
4260 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4261 way, but not yet received, when the check is performed.
4263 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4264 on TCP/IP sockets, because this apparently causes some broken clients to
4267 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4268 unchanged) from the Cygwin maintainer.
4270 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4271 was supported instead of testing for mbx. Effectively a typo.
4273 15. The spa authenticator server code was not checking that the input it
4274 received was valid base64.
4276 16. The debug output line for the "set" modifier in ACLs was not showing the
4277 name of the variable that was being set.
4279 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4281 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4282 lookup on its own. Something else was needed to trigger the lookup. For
4283 example, a match in host_lookup or the need for a host name in a host list.
4284 Now, if $sender_host_name is referenced and the host name has not yet been
4285 looked up, a lookup is performed. If the lookup fails, the variable remains
4286 empty, and $host_lookup_failed is set to "1".
4288 19. Added "eqi" as a case-independent comparison operator.
4290 20. The saslauthd authentication condition could segfault if neither service
4291 nor realm was specified.
4293 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4294 error message that was logged was misleading, and incoming SMTP
4295 connections were dropped. The message is now more accurate, and temporary
4296 errors are given to SMTP connections.
4298 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4299 (except RSET) with a 421 error, until QUIT is received. However, it was
4300 failing to send a response to QUIT.
4302 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4303 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4304 ACL is now run after the helo_try_verify_hosts code.
4306 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4307 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4308 case-independent, so other case variants are also recognized). Apparently
4309 some systems use these upper case variants.
4311 25. If more than two messages were waiting for the same host, and a transport
4312 filter was specified for the transport, Exim sent two messages over the
4313 same TCP/IP connection, and then failed with "socket operation on non-
4314 socket" when it tried to send the third.
4316 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4318 27. The extern definition of crypt16() in expand.c was not being excluded when
4319 the OS had its own crypt16() function.
4321 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4322 as a preferred synonym for return_size_limit, both as an option and as an
4325 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4327 30. Changed the default configuration ACL to relax the local part checking rule
4328 for addresses that are not in any local domains. For these addresses,
4329 slashes and pipe symbols are allowed within local parts, but the sequence
4330 /../ is explicitly forbidden.
4332 31. SPA server authentication was not clearing the challenge buffer before
4335 32. log_message in a "warn" ACL statement was writing to the reject log as
4336 well as to the main log, which contradicts the documentation and doesn't
4337 seem right (because no rejection is happening). So I have stopped it.
4339 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4340 However, I am unable to do any testing of this.
4342 34. Fixed an infelicity in the appendfile transport. When checking directories
4343 for a mailbox, to see if any needed to be created, it was accidentally
4344 using path names with one or more superfluous leading slashes; tracing
4345 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4347 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4348 contents are added to the log line that is written for every discarded
4349 recipient. (Previously a log_message setting was ignored.)
4351 36. The ${quote: operator now quotes the string if it is empty.
4353 37. The install script runs exim in order to find its version number. If for
4354 some reason other than non-existence or emptiness, which it checks, it
4355 could not run './exim', it was installing it with an empty version number,
4356 i.e. as "exim-". This error state is now caught, and the installation is
4359 38. An argument was missing from the function that creates an error message
4360 when Exim fails to connect to the socket for saslauthd authentication.
4361 This could cause Exim to crash, or give a corrupted message.
4363 39. Added isip, isip4, and isip6 to ${if conditions.
4365 40. The ACL variables $acl_xx are now saved with the message, and can be
4366 accessed later in routers, transports, and filters.
4368 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4369 strings in the file are not string-expanded.
4371 42. If a MAIL command specified a SIZE value that was too large to fit into an
4372 int variable, the check against message_size_limit failed. Such values are
4373 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4374 day this will have to be increased, but I don't think I want to be around
4375 when emails are that large.
4382 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4383 iconv() is not standard in FreeBSD.
4385 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4386 IPv6 enabled. The observed symptom was a segmentation fault on return from
4387 the function os_common_find_running_interfaces() in src/os.c.
4389 3. In the check_special_case() function in daemon.c I had used "errno" as an
4390 argument name, which causes warnings on some systems. This was basically a
4391 typo, since it was named "eno" in the comments!
4393 4. The code that waits for the clock to tick (at a resolution of some fraction
4394 of a second) so as to ensure message-id uniqueness was always waiting for
4395 at least one whole tick, when it could have waited for less. [This is
4396 almost certainly not relevant at current processor speeds, where it is
4397 unlikely to ever wait at all. But we try to future-proof.]
4399 5. The function that sleeps for a time interval that includes fractions of a
4400 second contained a race. It did not block SIGALRM between setting the
4401 timer, and suspending (a couple of lines later). If the interval was short
4402 and the sigsuspend() was delayed until after it had expired, the suspension
4403 never ended. On busy systems this could lead to processes getting stuck for
4406 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4407 process, before it forks any delivery processes. The open lookup caching
4408 mechanism meant that the open file or database connection was passed into
4409 the delivery process. The problem was that delivery processes always tidy
4410 up cached lookup data. This could cause a problem for the next delivery
4411 process started by the queue runner, because the external queue runner
4412 process does not know about the closure. So the next delivery process
4413 still has data in the lookup cache. In the case of a file lookup, there was
4414 no problem because closing a file descriptor in a subprocess doesn't affect
4415 the parent. However, if the lookup was caching a connection to a database,
4416 the connection was closed, and the second delivery process was likely to
4417 see errors such as "PGSQL: query failed: server closed the connection
4418 unexpectedly". The problem has been fixed by closing all cached lookups
4419 in a queue runner before running a delivery process.
4421 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4422 seem to have the "const" qualifier which it has on other OS. I've
4425 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4426 *of the same type* (client or server) with the same public name that an
4427 error should be diagnosed.
4429 9. When Exim looked up a host name for an IP address, but failed to find the
4430 original IP address when looking up the host name (a safety check), it
4431 output the message "<ip address> does not match any IP for NULL", which was
4432 confusing, to say the least. The bug was that the host name should have
4433 appeared instead of "NULL".
4435 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4436 user that is built into the binary, and the -C or -D options is used, root
4437 privilege is dropped before the configuration file is read. In addition,
4438 logging is switched to stderr instead of the normal log files. If the
4439 configuration then re-defines the Exim user, the unprivileged environment
4440 is probably not what is expected, so Exim logs a panic warning message (but
4443 However, if deliver_drop_privilege is set, the unprivileged state may well
4444 be exactly what is intended, so the warning has been cut out in that case,
4445 and Exim is allowed to try to write to its normal log files.
4451 1. smtp_return_error_details was not giving details for temporary sender
4452 or receiver verification errors.
4454 2. Diagnose a configuration error if two authenticators have the same public
4457 3. Exim used not to create the message log file for a message until the first
4458 delivery attempt. This could be confusing when incoming messages were held
4459 for policy or load reasons. The message log file is now created at the time
4460 the message is received, and an initial "Received" line is written to it.
4462 4. The automatically generated man page for command line options had a minor
4463 bug that caused no ill effects; however, a more serious problem was that
4464 the procedure for building the man page automatically didn't always
4465 operate. Consequently, release 4.20 contains an out-of-date version. This
4466 shouldn't happen again.
4468 5. When building Exim with embedded Perl support, the script that builds the
4469 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4470 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4472 6. The freeze_tell option was not being used for messages that were frozen on
4473 arrival, either by an ACL or by local_scan().
4475 7. Added the smtp_incomplete_transaction log selector.
4477 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4478 was accepting AUTH without a new EHLO.
4480 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4481 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4484 10. Logging of TCP/IP connections (when configured) now happens in the main
4485 daemon process instead of the child process, so that the TCP/IP connection
4486 count is more accurate (but it can never be perfect).
4488 11. The use of "drop" in a nested ACL was not being handled correctly in the
4489 outer ACL. Now, if condition failure induced by the nested "drop" causes
4490 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4491 or "require"), the connection is dropped.
4493 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4494 that yield "discard" can now be used with an "accept" or a "discard" verb,
4495 but an error is generated for any others (because I can't see a useful way
4496 to define what should happen).
4498 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4499 are now processed in the same way as lines in the Exim configuration file.
4500 In particular, continuation lines are supported.
4502 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4504 15. Added -ti meaning -t -i.
4506 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4507 domains, and warn by logging if others are found.
4509 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4510 returned from GIFCONF when Exim is trying to find the list of interfaces on
4511 a host. The code in os.c has been modified to copy each ifreq to an aligned
4512 structure in all cases.
4514 Also, in some cases, the returned ifreq's were being copied to a 'struct
4515 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4516 means the last couple of bytes of an IPv6 address could be chopped if the
4517 ifreq contained only a normal sockaddr (14 bytes storage).
4519 18. Named domain lists were not supported in the hosts_treat_as_local option.
4520 An entry such as +xxxx was not recognized, and was treated as a literal
4523 19. Ensure that header lines added by a DATA ACL are included in the reject log
4524 if the ACL subsequently rejects the message.
4526 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4527 MD5 (which is deprecated).
4529 21. When testing a filter file using -bf, Exim was writing a message when it
4530 took the sender from a "From " line in the message, but it was not doing so
4531 when it took $return_path from a Return-Path: header line. It now does.
4533 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4534 with a valid header line field name (a series of printing characters
4535 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4537 23. Changed "disc" in the source to "disk" to conform to the documentation and
4538 the book and for uniformity.
4540 24. Ignore Sendmail's -Ooption=value command line item.
4542 25. When execve() failed while trying to run a command in a pipe transport,
4543 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4544 could be confused with a return value of 69 from the command itself. This
4545 has been changed to 127, the value the shell returns if it is asked to run
4546 a non-existent command. The wording for the related log line suggests a
4547 non-existent command as the problem.
4549 26. If received_header_text expands to an empty string, do not add a Received:
4550 header line to the message. (Well, it adds a token one on the spool, but
4551 marks it "old" so that it doesn't get used or transmitted.)
4553 27. Installed eximstats 1.28 (addition of -nt option).
4555 28. There was no check for failure on the call to getsockname() in the daemon
4556 code. This can fail if there is a shortage of resources on the system, with
4557 ENOMEM, for example. A temporary error is now given on failure.
4559 29. Contrary to the C standard, it seems that in some environments, the
4560 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4561 program. Exim now does this explicitly; it affects the formatting of
4562 timestamps using strftime().
4564 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4565 complaints. I've now initialized all the variables, to avoid this.
4567 32. Header lines added by a system filter were not being "seen" during
4568 transport-time rewrites.
4570 33. The info_callback() function passed to OpenSSL is set up with type void
4571 (*)(SSL *, int, int), as described somewhere. However, when calling the
4572 function (actually a macro) that sets it up, the type void(*)() is
4573 expected. I've put in a cast to prevent warnings from picky compilers.
4575 34. If a DNS black list lookup found a CNAME record, but there were no A
4576 records associated with the domain it pointed at, Exim crashed.
4578 35. If a DNS black list lookup returned more than one A record, Exim ignored
4579 all but the first. It now scans all returned addresses if a particular IP
4580 value is being sought. In this situation, the contents of the
4581 $dnslist_value variable are a list of all the addresses, separated by a
4584 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4585 to accept a host name and all its aliases if the forward lookup for any of
4586 them yielded the IP address of the incoming connection. Now it accepts only
4587 those names whose forward lookup yields the correct IP address. Any other
4588 names are discarded. This closes a loophole whereby a rogue DNS
4589 administrator could create reverse DNS records to break through a
4590 wildcarded host restriction in an ACL.
4592 37. If a user filter or a system filter that ran in a subprocess used any of
4593 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4594 the wrong values were passed to the pipe command ($thisaddress had the
4595 value of $0, $0 had the value of $1, etc). This bug was introduced by
4596 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4598 38. Improved the line breaking for long SMTP error messages from ACLs.
4599 Previously, if there was no break point between 40 and 75 characters, Exim
4600 left the rest of the message alone. Two changes have been made: (a) I've
4601 reduced the minimum length to 35 characters; (b) if it can't find a break
4602 point between 35 and 75 characters, it looks ahead and uses the first one
4603 that it finds. This may give the occasional overlong line, but at least the
4604 remaining text gets split now.
4606 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4607 file descriptors might be low, and that setting 1000 would always raise it.
4608 It turns out that in some environments, the limit is already over 1000 and
4609 that lowering it causes trouble. So now Exim takes care not to decrease it.
4611 40. When delivering a message, the value of $return_path is set to $sender_
4612 address at the start of routing (routers may change the value). By an
4613 oversight, this default was not being set up when an address was tested by
4614 -bt or -bv, which affected the outcome if any router or filter referred to
4617 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4618 log only when "message" or "log_message" are set. However, if one of the
4619 conditions was an address verification, or a call to a nested ACL, the
4620 messages generated by the underlying test were being passed through. This
4621 no longer happens. The underlying message is available in $acl_verify_
4622 message for both "message" and "log_message" expansions, so it can be
4623 passed through if needed.
4625 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4626 new expansion $bh_ to give the encoded byte string without charset
4627 translation. Translation happens only if iconv() is available; HAVE_ICONV
4628 indicates this at build time. HEADERS_CHARSET gives the charset to
4629 translate to; headers_charset can change it in the configuration, and
4630 "headers charset" can change it in an individual filter file.
4632 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4633 that creates RFC 2047 encoded "words" labels them as that charset instead
4634 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4635 expansion operator; (ii) when Exim creates a From: line for a local
4636 message; (iii) when a header line is rewritten to include a "phrase" part.
4638 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4639 buggy, causing it to skip the first lines of messages whose message ID
4640 ended in 'D'. This would not have bitten before Exim release 4.14, because
4641 message IDs were unlikely to end in 'D' before then. The effect was to have
4642 incorrect size information for certain domains.
4644 45. #include "config.h" was missing at the start of the crypt16.c module. This
4645 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4648 46. If there was a timeout during a "random" callout check, Exim treated it as
4649 a failure of the random address, and carried on sending RSET and the real
4650 address. If the delay was just some slowness somewhere, the response to the
4651 original RCPT would be taken as a response to RSET and so on, causing
4652 mayhem of various kinds.
4654 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4655 when I implemented it. It didn't allow for the fact that some option values
4656 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4657 the right test for positive values.
4659 48. Domain names in DNS records are case-independent. Exim always looks them up
4660 in lower case. Some resolvers return domain names in exactly the case they
4661 appear in the zone file, that is, they may contain uppercase letters. Not
4662 all resolvers do this - some return always lower case. Exim was treating a
4663 change of case by a resolver as a change of domain, similar to a widening
4664 of a domain abbreviation. This triggered its re-routing code and so it was
4665 trying to route what was effectively the same domain again. This normally
4666 caused routing to fail (because the router wouldn't handle the domain
4667 twice). Now Exim checks for this case specially, and just changes the
4668 casing of the domain that it ultimately uses when it transmits the message
4671 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4674 50. If a filter generated a file delivery with a non-absolute name (possible if
4675 no home directory exists for the router), the forbid_file option was not
4678 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4679 the existing equality matching.
4681 52. Exim was using ints instead of ino_t variables in some places where it was
4682 dealing with inode numbers.
4684 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4685 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4686 called TMPDIR, and if it finds it is different, it changes its value.
4688 54. The smtp_printf() function is now made available to local_scan() so
4689 additional output lines can be written before returning. There is also an
4690 smtp_fflush() function to enable the detection of a dropped connection.
4691 The variables smtp_input and smtp_batched_input are exported to
4694 55. Changed the default runtime configuration: the message "Unknown user"
4695 has been removed from the ACL, and instead placed on the localuser router,
4696 using the cannot_route_message feature. This means that any verification
4697 failures that generate their own messages won't get overridden. Similarly,
4698 the "Unrouteable address" message that was in the ACL for unverifiable
4699 relay addresses has also been removed.
4701 56. Added hosts_avoid_esmtp to the smtp transport.
4703 57. The exicyclog script was not checking for the esoteric option
4704 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4705 will work only if exicyclog is run under the appropriate euid.
4707 58. Following a discussion on the list, the rules by which Exim recognises line
4708 endings on incoming messages have been changed. The -dropcr and drop_cr
4709 options are now no-ops, retained only for backwards compatibility. The
4710 following line terminators are recognized: LF CRLF CR. However, special
4711 processing applies to CR:
4713 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4714 nor a local message in the state where . is a terminator.
4716 (ii) If a bare CR is encountered in a header line, an extra space is added
4717 after the line terminator so as not to end the header. The reasoning
4718 behind this is that bare CRs in header lines are most likely either
4719 to be mistakes, or people trying to play silly games.
4721 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4722 was being incorrectly given as 18 bytes larger than it should have been.
4723 This is a VOB (very old bug).
4725 60. This may never have affected anything current, but just in case it has:
4726 When the local host is found other than at the start of a list of hosts,
4727 the local host, those with the same MX, and any that follow, are discarded.
4728 When the list in question was part of a longer list of hosts, the following
4729 hosts (not currently being processed) were also being discarded. This no
4730 longer happens. I'm not sure if this situation could ever has previously
4733 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4736 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4737 header. This is something that is recommended in a new Internet Draft, and
4738 is something that is documented as being done by Sendmail. There are two
4739 possible values. For messages generated by the autoreply transport, Exim
4742 Auto-Submitted: auto-replied
4744 whereas for all other generated messages (e.g. bounces) it adds
4746 Auto-Submitted: auto-generated
4748 63. The "personal" condition in filters now includes a test for the
4749 Auto-Submitted: header. If it contains the string "auto-" the message it
4750 not considered personal.
4752 64. Added rcpt_include_affixes as a generic transport option.
4754 65. Added queue_only_override (default true).
4756 66. Added the syslog_duplication option.
4758 67. If what should have been the first header line of a message consisted of
4759 a space followed by a colon, Exim was mis-interpreting it as a header line.
4760 It isn't of course - it is syntactically invalid and should therefore be
4761 treated as the start of the message body. The misbehaviour could have
4762 caused a number of strange effects, including loss of data in subsequent
4763 header lines, and spool format errors.
4765 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4766 client host had authenticated. This control can now be exercised by an ACL
4767 for more flexibility.
4769 69. By default, callouts do not happen when testing with -bh. There is now a
4770 variant, -bhc, which does actually run the callout code, including
4771 consulting and updating the callout cache.
4773 70. Added support for saslauthd authentication, courtesy of Alexander
4776 71. If statvfs() failed on the spool or log directories while checking their
4777 size for availability, Exim confusingly gave the error "space shortage".
4778 Furthermore, in debugging mode it crashed with a floating point exception.
4779 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4780 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4781 serious problem, Exim now writes to the main and panic logs when this
4782 happens, with details of the failure. It then refuses to accept the
4783 incoming message, giving the message "spool directory problem" or "log
4784 directory problem" with a 421 code for SMTP messages.
4786 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4787 0, 1, and 2 exist, because some OS complain for execs without them (see
4788 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4789 descriptors. However, the code omitted to check that the open succeeded,
4790 causing mysterious errors if for some reason the permissions on /dev/null
4791 got screwed. Now Exim writes a message to the main and panic logs, and
4792 bombs out if it can't open /dev/null.
4794 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4795 interact so that it is all more flexible. It is supposed to remain
4796 backwards compatible. Also added extra_local_interfaces.
4798 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4799 to bomb out with an assertion failure - to the client this appears as a
4800 connection drop. This problem occurs in the part of the code that was taken
4801 from the Samba project. Fortunately, the assertion is in a very simple
4802 function, so I have fixed this by reproducing the function inline in the
4803 one place where it is called, and arranging for authentication to fail
4804 instead of killing the process with assert().
4806 75. The SPA client code was not working when the server requested OEM rather
4807 than Unicode encoding.
4809 76. Added code to make require_files with a specific uid setting more usable in
4810 the case where statting the file as root fails - usually a non-root-mounted
4811 NFS file system. When this happens and the failure is EACCES, Exim now
4812 forks a subprocess and does the per-uid checking as the relevant uid.
4814 77. Added process_log_path.
4816 78. If log_file_path was not explicitly set, a setting of check_log_space or
4817 check_log_inodes was ignored.
4819 79. If a space check for the spool or log partitions fails, the incident is now
4820 logged. Of course, in the latter case the data may get lost...
4822 80. Added the %p formatting code to string_format() so that it can be used to
4823 print addresses in debug_print(). Adjusted all the address printing in the
4824 debugging in store.c to use %p rather than %d.
4826 81. There was a concern that a line of code in smtp_in.c could overflow a
4827 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4828 initially expressed, the concern was not well-founded, because trailing
4829 spaces are removed early. However, if the trailing spaces were followed by
4830 a NULL, they did not get removed, so the overflow was possible. Two fixes
4833 (a) I re-wrote the offending code in a cleaner fashion.
4834 (b) If an incoming SMTP command contains a NULL character, it is rejected
4837 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4838 runs initgroups(), so that if the Exim user is in any additional groups,
4839 they will be used during message reception.
4845 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob