1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.536 2007/11/12 13:02:19 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Add preliminary DKIM support. Currently requires a forked version of
10 ALT-N's libdkim that I have put here:
11 http://duncanthrax.net/exim-experimental/
13 Note to Michael Haardt: I had to rename some vars in sieve.c. They
14 were called 'true' and it seems that C99 defines that as a reserved
15 keyword to be used with 'bool' variable types. That means you could
16 not include C99-style headers which use bools without triggering
17 build errors in sieve.c.
19 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
20 as mailq or other aliases. Changed the --help handling significantly
21 to do whats expected. exim_usage() emits usage/help information.
23 SC/01 Added the -bylocaldomain option to eximstats.
25 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
27 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
29 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
35 PH/01 Another patch from the Sieve maintainer.
37 PH/02 When an IPv6 address is converted to a string for single-key lookup
38 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
39 dots are used instead of colons so that keys in lsearch files need not
40 contain colons. This was done some time before quoting was made available
41 in lsearch files. However, iplsearch files do require colons in IPv6 keys
42 (notated using the quote facility) so as to distinguish them from IPv4
43 keys. This meant that lookups for IP addresses in host lists did not work
44 for iplsearch lookups.
46 This has been fixed by arranging for IPv6 addresses to be expressed with
47 colons if the lookup type is iplsearch. This is not incompatible, because
48 previously such lookups could never work.
50 The situation is now rather anomolous, since one *can* have colons in
51 ordinary lsearch keys. However, making the change in all cases is
52 incompatible and would probably break a number of configurations.
54 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
57 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
58 conversion specification without a maximum field width, thereby enabling
59 a rogue spamd server to cause a buffer overflow. While nobody in their
60 right mind would setup Exim to query an untrusted spamd server, an
61 attacker that gains access to a server running spamd could potentially
62 exploit this vulnerability to run arbitrary code as the Exim user.
64 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
65 $primary_hostname instead of what libspf2 thinks the hosts name is.
67 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
68 a directory entry by the name of the lookup key. Previously, if a
69 symlink pointed to a non-existing file or a file in a directory that
70 Exim lacked permissions to read, a lookup for a key matching that
71 symlink would fail. Now it is enough that a matching directory entry
72 exists, symlink or not. (Bugzilla 503.)
74 PH/03 The body_linecount and body_zerocount variables are now exported in the
77 PH/04 Added the $dnslist_matched variable.
79 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
80 This means they are set thereafter only if the connection becomes
83 PH/06 Added the client_condition to authenticators so that some can be skipped
84 by clients under certain conditions.
86 PH/07 The error message for a badly-placed control=no_multiline_responses left
87 "_responses" off the end of the name.
89 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
91 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
92 (without spaces) instead of just copying the configuration text.
94 PH/10 Added the /noupdate option to the ratelimit ACL condition.
96 PH/11 Added $max_received_linelength.
98 PH/12 Added +ignore_defer and +include_defer to host lists.
100 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
101 way in which PCRE > 7.0 is built.
103 PH/14 Implemented queue_only_load_latch.
105 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
106 MAIL command. The effect was to mangle the value on 64-bit systems.
108 PH/16 Another patch from the Sieve maintainer.
110 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
112 PH/18 If a system quota error occurred while trying to create the file for
113 a maildir delivery, the message "Mailbox is full" was not appended to the
114 bounce if the delivery eventually timed out. Change 4.67/27 below applied
115 only to a quota excession during the actual writing of the file.
117 PH/19 It seems that peer DN values may contain newlines (and other non-printing
118 characters?) which causes problems in log lines. The DN values are now
119 passed through string_printing() before being added to log lines.
121 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
122 and InterBase are left for another time.)
124 PH/21 Added message_body_newlines option.
126 PH/22 Guard against possible overflow in moan_check_errorcopy().
128 PH/23 POSIX allows open() to be a macro; guard against that.
130 PH/24 If the recipient of an error message contained an @ in the local part
131 (suitably quoted, of course), incorrect values were put in $domain and
132 $local_part during the evaluation of errors_copy.
138 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
139 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
142 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
143 issue a MAIL command.
145 PH/02 In an ACL statement such as
147 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
149 if a client was not listed at all, or was listed with a value other than
150 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
151 the condition was not true (as it should be), so access was not denied.
152 The bug was that the ! inversion was incorrectly passed on to the second
153 item. This has been fixed.
155 PH/03 Added additional dnslists conditions == and =& which are different from
156 = and & when the dns lookup returns more than one IP address.
158 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
159 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
161 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
162 FSYNC, which compiles an option called disable_fsync that allows for
163 bypassing fsync(). The documentation is heavily laced with warnings.
165 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
167 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
168 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
169 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
170 including adding "make clean"; (3) Added -fPIC when compiling the test
171 dynamically loaded module, to get rid of a warning.
173 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
174 message fails, move_frozen_messages = true and ignore_bounce_errors_after
175 = 0s. The bug is otherwise harmless.
177 PH/07 There was a bug in the dovecot authenticator such that the value of
178 $auth1 could be overwritten, and so not correctly preserved, after a
179 successful authentication. This usually meant that the value preserved by
180 the server_setid option was incorrect.
182 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
184 PH/09 Installed PCRE release 7.0.
186 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
187 run for batched SMTP input. It is now run at the start of every message
188 in the batch. While fixing this I discovered that the process information
189 (output by running exiwhat) was not always getting set for -bs and -bS
190 input. This is fixed, and it now also says "batched" for BSMTP.
192 PH/11 Added control=no_pipelining.
194 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
195 patch, slightly modified), and move the expansion of helo_data till after
196 the connection is made in the smtp transport (so it can use these
199 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
201 PH/14 Added log_selector = +pid.
203 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
205 PH/16 Add ${if forany and ${if forall.
207 PH/17 Added dsn_from option to vary the From: line in DSNs.
209 PH/18 Flush SMTP output before performing a callout, unless control =
210 no_callout_flush is set.
212 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
213 was true (the default) a successful delivery failed to delete the retry
214 item, thus causing premature timeout of the address. The bug is now
217 PH/20 Added hosts_avoid_pipelining to the smtp transport.
219 PH/21 Long custom messages for fakedefer and fakereject are now split up
220 into multiline reponses in the same way that messages for "deny" and
221 other ACL rejections are.
223 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
224 with slight modification.
226 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
227 draft, changing the syntax and factoring some duplicate code".
229 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
230 for deliveries of the second and subsequent messages over the same SMTP
233 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
234 ${reduce, with only minor "tidies".
236 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
238 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
239 expansion side effects.
241 PH/27 When a message times out after an over-quota error from an Exim-imposed
242 quota, the bounce message says "mailbox is full". This message was not
243 being given when it was a system quota that was exceeded. It now should
246 MH/03 Made $recipients available in local_scan(). local_scan() already has
247 better access to the recipient list through recipients_list[], but
248 $recipients can be useful in postmaster-provided expansion strings.
250 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
251 in the case of a MAIL command with additional options following the
252 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
253 were accidentally chopped off.
255 PH/29 SMTP synchronization checks are implemented when a command is read -
256 there is a check that no more input is waiting when there shouldn't be
257 any. However, for some commands, a delay in an ACL can mean that it is
258 some time before the response is written. In this time, more input might
259 arrive, invalidly. So now there are extra checks after an ACL has run for
260 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
261 pipelining has not been advertised.
263 PH/30 MH's patch to allow iscntrl() characters to be list separators.
265 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
266 returned in the SMTP response when smtp_return_error_details was false.
269 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
270 instead of the C I/O that was originally supplied, because problems were
273 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
274 Exim which did not show up earlier: it was assuming that a call to
275 SSL_CTX_set_info_callback() might give an error value. In fact, there is
276 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
277 was a macro that became an assignment, so it seemed to work. This has
278 changed to a proper function call with a void return, hence the compile
279 error. Exim's code has been fixed.
281 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
284 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
286 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
288 PH/37 If a message is not accepted after it has had an id assigned (e.g.
289 because it turns out to be too big or there is a timeout) there is no
290 "Completed" line in the log. When some messages of this type were
291 selected by exigrep, they were listed as "not completed". Others were
292 picked up by some special patterns. I have improved the selection
293 criteria to be more general.
295 PH/38 The host_find_failed option in the manualroute router can now be set
296 to "ignore", to completely ignore a host whose IP address cannot be
297 found. If all hosts are ignored, the behaviour is controlled by the new
298 host_all_ignored option.
300 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
301 homing or because of multiple MX records with /mx) generated more than
302 one IP address, and the following item turned out to be the local host,
303 all the secondary addresses of the first item were incorrectly removed
304 from the list, along with the local host and any following hosts (which
305 is what is supposed to happen).
307 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
308 whoever called Exim into the -H file. In the case of the daemon it was
309 behaving confusingly. When first started, it used values for whoever
310 started the daemon, but after a SIGHUP it used the Exim user (because it
311 calls itself on a restart). I have changed the code so that it now always
314 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
315 message are rejected with the same error (e.g. no authentication or bad
316 sender address), and a DATA command is nevertheless sent (as can happen
317 with PIPELINING or a stupid MUA), the error message that was given to the
318 RCPT commands is included in the rejection of the DATA command. This is
319 intended to be helpful for MUAs that show only the final error to their
322 PH/42 Another patch from the Sieve maintainer.
324 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
325 Eximstats - Fixed some broken HTML links and added missing column headers
327 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
328 columns for Rejects, Temp Rejects, Ham, and Spam rows.
330 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
332 PH/43 Yet another patch from the Sieve maintainer.
334 PH/44 I found a way to check for a TCP/IP connection going away before sending
335 the response to the final '.' that terminates a message, but only in the
336 case where the client has not sent further data following the '.'
337 (unfortunately, this is allowed). However, in many cases there won't be
338 any further data because there won't be any more messages to send. A call
339 to select() can be used: if it shows that the input is "ready", there is
340 either input waiting, or the socket has been closed. An attempt to read
341 the next input character can distinguish the two cases. Previously, Exim
342 would have sent an OK response which the client would never have see.
343 This could lead to message repetition. This fix should cure that, at
344 least in a lot of common cases.
346 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
347 advertised in response to EHLO.
353 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
354 fixed by 4.65/MH/01 (is this a record?) are fixed:
356 (i) An empty string was always treated as zero by the numeric comparison
357 operators. This behaviour has been restored.
359 (ii) It is documented that the numeric comparison operators always treat
360 their arguments as decimal numbers. This was broken in that numbers
361 starting with 0 were being interpreted as octal.
363 While fixing these problems I realized that there was another issue that
364 hadn't been noticed. Values of message_size_limit (both the global option
365 and the transport option) were treated as octal if they started with 0.
366 The documentation was vague. These values are now always treated as
367 decimal, and I will make that clear in the documentation.
373 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
374 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
377 MH/01 Don't check that the operands of numeric comparison operators are
378 integers when their expansion is in "skipping" mode (fixes bug
379 introduced by 4.64-PH/07).
381 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
382 child addresses, Exim now panics and dies. Previously, because the count
383 is held in a short int, deliveries were likely to be lost. As such a
384 large number of recipients for a single message is ridiculous
385 (performance will be very, very poor), I have chosen to impose a limit
386 rather than extend the field.
392 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
393 leftover -K file (the existence of which was triggered by #402).
394 While we were at it, introduced process PID as part of the -K
395 filename. This should rule out race conditions when creating
398 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
399 processing considerably. Previous code took too long for large mails,
400 triggering a timeout which in turn triggers #401.
402 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
403 in the DK code in transports.c. sendfile() is not really portable,
404 hence the _LINUX specificness.
406 TF/01 In the add_headers option to the mail command in an Exim filter,
407 there was a bug that Exim would claim a syntax error in any
408 header after the first one which had an odd number of characters
411 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
412 callout verification, Exim cached a "reject" for the entire domain. This
413 is correct for most verifications, but it is not correct for a recipient
414 verification with use_sender or use_postmaster set, because in that case
415 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
416 case of MAIL FROM:<> rejection from other early rejections (e.g.
417 rejection of HELO). When verifying a recipient using a non-null MAIL
418 address, the cache is ignored if it shows MAIL FROM:<> rejection.
419 Whatever the result of the callout, the value of the domain cache is
420 left unchanged (for any other kind of callout, getting as far as trying
421 RCPT means that the domain itself is ok).
423 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
426 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
427 manifest itself as EPIPE rather than ECONNECT. When tidying away a
428 session, the daemon ignores ECONNECT errors and logs others; it now
429 ignores EPIPE as well.
431 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
432 (quoted-printable decoding).
434 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
435 later the small subsequent patch to fix an introduced bug.
437 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
439 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
441 PH/08 An error is now given if message_size_limit is specified negative.
443 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
444 to be given (somewhat) arbitrary names.
446 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
449 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
450 miscellaneous code fixes
452 PH/10 Added the log_reject_target ACL modifier to specify where to log
455 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
456 hostname. This is wrong, because it relates to the incoming message (and
457 probably the interface on which it is arriving) and not to the outgoing
458 callout (which could be using a different interface). This has been
459 changed to use the value of the helo_data option from the smtp transport
460 instead - this is what is used when a message is actually being sent. If
461 there is no remote transport (possible with a router that sets up host
462 addresses), $smtp_active_hostname is used.
464 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
465 tweaks were necessary in order to get it to work (see also 21 below):
466 (a) The code assumed that strncpy() returns a negative number on buffer
467 overflow, which isn't the case. Replaced with Exim's string_format()
469 (b) There were several signed/unsigned issues. I just did the minimum
470 hacking in of casts. There is scope for a larger refactoring.
471 (c) The code used strcasecmp() which is not a standard C function.
472 Replaced with Exim's strcmpic() function.
473 (d) The code set only $1; it now sets $auth1 as well.
474 (e) A simple test gave the error "authentication client didn't specify
475 service in request". It would seem that Dovecot has changed its
476 interface. Fortunately there's a specification; I followed it and
477 changed what the client sends and it appears to be working now.
479 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
482 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
483 address A is aliased to B and C, where B exists and C does not. Without
484 -v the output is "A verified" because verification stops after a
485 successful redirection if more than one address is generated. However,
486 with -v the child addresses are also verified. Exim was outputting "A
487 failed to verify" and then showing the successful verification for C,
488 with its parentage. It now outputs "B failed to verify", showing B's
489 parentage before showing the successful verification of C.
491 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
492 look up a TXT record in a specific list after matching in a combined
495 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
496 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
497 they consult the DNS. I had assumed they would set it the way they
498 wanted; and indeed my experiments on Linux seem to show that in some
499 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
500 To be on the safe side, however, I have now made the interface to
501 host_find_byname() similar to host_find_bydns(), with an argument
502 containing the DNS resolver options. The host_find_byname() function now
503 sets these options at its start, just as host_find_bydns() does. The smtp
504 transport options dns_qualify_single and dns_search_parents are passed to
505 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
506 of host_find_byname() use the default settings of RES_DEFNAMES
507 (qualify_single) but not RES_DNSRCH (search_parents).
509 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
510 spool_read_header() do less string testing, by means of a preliminary
511 switch on the second character of optional "-foo" lines. (This is
512 overdue, caused by the large number of possibilities that now exist.
513 Originally there were few.) While I was there, I also converted the
514 str(n)cmp tests so they don't re-test the leading "-" and the first
515 character, in the hope this might squeeze out yet more improvement.
517 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
518 flag allowing group syntax was set by the header_syntax check but not
519 turned off, possible causing trouble later; (2) The flag was not being
520 set at all for the header_verify test, causing "group"-style headers to
521 be rejected. I have now set it in this case, and also caused header_
522 verify to ignore an empty address taken from a group. While doing this, I
523 came across some other cases where the code for allowing group syntax
524 while scanning a header line wasn't quite right (mostly, not resetting
525 the flag correctly in the right place). These bugs could have caused
526 trouble for malformed header lines. I hope it is now all correct.
528 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
529 with the "reply" argument non-NULL. The code, however (which originally
530 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
531 but it didn't always do it. This confused somebody who was copying the
532 code for some other use. I have removed all the tests.
534 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
535 feature that was used to support insecure browsers during the U.S. crypto
536 embargo. It requires special client support, and Exim is probably the
537 only MTA that supported it -- and would never use it because real RSA is
538 always available. This code has been removed, because it had the bad
539 effect of slowing Exim down by computing (never used) parameters for the
540 RSA_EXPORT functionality.
542 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
543 authenticator to fail if there's a tab character in the incoming data
544 (there should never be unless someone is messing about, as it's supposed
545 to be base64-encoded). Also added, on Timo's advice, the "secured" option
546 if the connection is using TLS or if the remote IP is the same as the
547 local IP, and the "valid-client-cert option" if a client certificate has
550 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
551 authenticators. This can be used for authorization after authentication
552 succeeds. (In the case of plaintext, it servers for both authentication
555 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
556 if any retry times were supplied.
558 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
559 connection, where there is no client IP address to check. In this
560 situation, the verify now always succeeds.
562 PH/25 Applied John Jetmore's -Mset patch.
564 PH/26 Added -bem to be like -Mset, but loading a message from a file.
566 PH/27 In a string expansion for a processed (not raw) header when multiple
567 headers of the same name were present, leading whitespace was being
568 removed from all of them, but trailing whitespace was being removed only
569 from the last one. Now trailing whitespace is removed from each header
570 before concatenation. Completely empty headers in a concatenation (as
573 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
574 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
576 PH/29 [Removed. This was a change that I later backed out, and forgot to
577 correct the ChangeLog entry (that I had efficiently created) before
578 committing the later change.]
580 PH/30 Exim was sometimes attempting to deliver messages that had suffered
581 address errors (4xx response to RCPT) over the same connection as other
582 messages routed to the same hosts. Such deliveries are always "forced",
583 so retry times are not inspected. This resulted in far too many retries
584 for the affected addresses. The effect occurred only when there were more
585 hosts than the hosts_max_try setting in the smtp transport when it had
586 the 4xx errors. Those hosts that it had tried were not added to the list
587 of hosts for which the message was waiting, so if all were tried, there
588 was no problem. Two fixes have been applied:
590 (i) If there are any address or message errors in an SMTP delivery, none
591 of the hosts (tried or untried) are now added to the list of hosts
592 for which the message is waiting, so the message should not be a
593 candidate for sending over the same connection that was used for a
594 successful delivery of some other message. This seems entirely
595 reasonable: after all the message is NOT "waiting for some host".
596 This is so "obvious" that I'm not sure why it wasn't done
597 previously. Hope I haven't missed anything, but it can't do any
598 harm, as the worst effect is to miss an optimization.
600 (ii) If, despite (i), such a delivery is accidentally attempted, the
601 routing retry time is respected, so at least it doesn't keep
602 hammering the server.
604 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
605 in ${readsocket because some servers need this prod.
607 PH/32 Added some extra debug output when updating a wait-xxx database.
609 PH/33 The hint "could be header name not terminated by colon", which has been
610 given for certain expansion errors for a long time, was not being given
611 for the ${if def:h_colon_omitted{... case.
613 PH/34 The spec says: "With one important exception, whenever a domain list is
614 being scanned, $domain contains the subject domain." There was at least
615 one case where this was not true.
617 PH/35 The error "getsockname() failed: connection reset by peer" was being
618 written to the panic log as well as the main log, but it isn't really
619 panic-worthy as it just means the connection died rather early on. I have
620 removed the panic log writing for the ECONNRESET error when getsockname()
623 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
624 runs only) independently of the message's sender address. This meant
625 that, if the 4xx error was in fact related to the sender, a different
626 message to the same recipient with a different sender could confuse
627 things. In particualar, this can happen when sending to a greylisting
628 server, but other circumstances could also provoke similar problems.
629 I have changed the default so that the retry time for these errors is now
630 based a combination of the sender and recipient addresses. This change
631 can be overridden by setting address_retry_include_sender=false in the
634 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
635 remote server are returned as part of bounce messages. This was not
636 happening for LMTP over a pipe (the lmtp transport), but now it is the
637 same for both kinds of LMTP.
639 PH/38 Despite being documented as not happening, Exim was rewriting addresses
640 in header lines that were in fact CNAMEs. This is no longer the case.
642 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
643 and queue runs started by the daemon processed all messages. This has
644 been fixed so that -R and -S can now usefully be given with -q<time>.
646 PH/40 Import PCRE release 6.7 (fixes some bugs).
648 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
650 PH/42 Give an error if -q is specified more than once.
652 PH/43 Renamed the variables $interface_address and $interface_port as
653 $received_ip_address and $received_port, to make it clear that these
654 values apply to message reception, and not to the outgoing interface when
655 a message is delivered. (The old names remain recognized, of course.)
657 PH/44 There was no timeout on the connect() call when using a Unix domain
658 socket in the ${readsocket expansion. There now is.
660 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
661 be meaningful with "accept".
663 SC/01 Eximstats V1.43
664 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
666 SC/02 Eximstats V1.44
667 Use a glob alias rather than an array ref in the generated
668 parser. This improves both readability and performance.
670 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
671 Collect SpamAssassin and rejection statistics.
672 Don't display local sender or destination tables unless
673 there is data to show.
674 Added average volumes into the top table text output.
676 SC/04 Eximstats V1.46
677 Collect data on the number of addresses (recipients)
678 as well as the number of messages.
680 SC/05 Eximstats V1.47
681 Added 'Message too big' to the list of mail rejection
682 reasons (thanks to Marco Gaiarin).
684 SC/06 Eximstats V1.48
685 Mainlog lines which have GMT offsets and are too short to
686 have a flag are now skipped.
688 SC/07 Eximstats V1.49 (Alain Williams)
689 Added the -emptyok flag.
691 SC/08 Eximstats V1.50
692 Fixes for obtaining the IP address from reject messages.
694 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
695 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
696 whitesspace changes from 4.64-PH/27
698 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
701 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
702 are found, allow negative numbers in numeric criteria)
704 JJ/06 exipick.20061117.2, added new $message_body_missing variable
706 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
707 to match changes made in 4.64-PH/43
709 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
711 PH/47 Put in an explicit test for a DNS lookup of an address record where the
712 "domain" is actually an IP address, and force a failure. This locks out
713 those revolvers/nameservers that support "A-for-A" lookups, in
714 contravention of the specifications.
716 PH/48 When a host name was looked up from an IP address, and the subsequent
717 forward lookup of the name timed out, the host name was left in
718 $sender_host_name, contrary to the specification.
720 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
721 restricted to single-key lookups, Exim was not diagnosing an error if
722 * or *@ was used with a query-style lookup.
724 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
726 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
727 long ago, but noone interested enough thought of it. Let's just say that
728 the "1.1" means that there are some new functions that weren't there at
729 some point in the past.
731 PH/51 Error processing for expansion failure of helo_data from an smtp
732 transport during callout processing was broken.
734 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
735 tested/used via the -bh/-bhc/-bs options.
737 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
738 bug, fixed in subsequent PCRE releases).
740 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
741 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
743 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
749 SC/01 Use a glob alias rather than an array ref in eximstats generated
750 parser. This improves both readability and performance.
752 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
753 Don't display local sender or destination tables in eximstats unless
754 there is data to show.
755 Added average volumes into the eximstats top table text output.
757 SC/03 Collect data on the number of addresses (recipients) as well
758 as the number of messages in eximstats.
760 TF/01 Correct an error in the documentation for the redirect router. Exim
761 does (usually) call initgroups() when daemonizing.
763 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
764 with consistent privilege compared to when running as a daemon.
766 TF/03 Note in the spec that $authenticated_id is not set for local
767 submissions from trusted users.
769 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
770 Thanks to Dean Brooks <dean@iglou.com> for the patch.
772 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
773 by adding some example configuration directives to the default
774 configuration file. A little bit of work is required to uncomment the
775 directives and define how usernames and passwords are checked, but
776 there is now a framework to start from.
778 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
779 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
780 without this. I don't know how relevant this is to other LDAP libraries.
782 PH/02 Add the verb name to the "unknown ACL verb" error.
784 PH/03 Magnus Holmgren's patch for filter_prepend_home.
786 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
788 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
789 directory not expanded when it should be if an expanded home directory
790 was set for the address (which is overridden by the transport).
792 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
795 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
796 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
797 because it is too late at that time, and has no effect.
799 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
800 security issue with \' (bugzilla #107). I could not use the
801 PQescapeStringConn() function, because it needs a PGconn value as one of
804 PH/08 When testing addresses using -bt, indicate those final addresses that
805 are duplicates that would not cause an additional delivery. At least one
806 person was confused, thinking that -bt output corresponded to deliveries.
807 (Suppressing duplicates isn't a good idea as you lose the information
808 about possibly different redirections that led to the duplicates.)
810 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
811 systems where poll() doesn't work, in particular OS X.
813 PH/10 Added more information to debugging output for retry time not reached.
815 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
816 operations in malware.c.
818 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
821 PH/13 If write_rejectlog was set false when logging was sent to syslog with
822 syslog_duplication set false, log lines that would normally be written
823 both the the main log and to the reject log were not written to syslog at
826 PH/14 In the default configuration, change the use of "message" in ACL warn
827 statements to "add_header".
829 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
830 not followed by a command (e.g. "seen endif").
832 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
833 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
836 PH/17 Added extra conditions to the default value of delay_warning_condition
840 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
841 { match{$h_precedence:}{(?i)bulk|list|junk} } \
842 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
845 The Auto-Submitted: and various List- headers are standardised, whereas I
846 don't think Precedence: ever was.
848 PH/18 Refactored debugging code in route_finduser() to show more information,
849 in particular, the error code if getpwnam() issues one.
851 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
852 This is apparently needed in addition to the PH/07 change above to avoid
853 any possible encoding problems.
855 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
856 but not after initializing Perl.
858 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
859 output them only if debugging. By default they are written stderr,
860 apparently, which is not desirable.
862 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
865 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
868 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
870 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
871 authenticated or an ident call has been made. Suppress the default
872 values for $authenticated_id and $authenticated_sender (but permit -oMai
873 and -oMas) when testing with -bh.
875 PH/24 Re-jigged the order of the tests in the default configuration so that the
876 tests for valid domains and recipients precede the DNS black list and CSA
877 tests, on the grounds that those ones are more expensive.
879 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
880 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
881 This bug exists in every version of Exim that I still have, right back to
884 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
885 However, an attempt to turn on case-sensitivity in a regex key by
886 including (?-i) didn't work because the subject string was already
887 lowercased, and the effects were non-intuitive. It turns out that a
888 one-line patch can be used to allow (?-i) to work as expected.
894 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
895 other effects) broke the use of negated acl sub-conditions.
897 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
900 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
901 "Deny" causes Exim to reject the incoming connection with a 554 error.
902 Unfortunately, if there is a major crisis, such as a disk failure,
903 tcp-wrappers gives "deny", whereas what one would like would be some
904 kind of temporary error. A kludge has been added to help with this.
905 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
906 554 error is used if errno is still zero or contains ENOENT (which occurs
907 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
910 PH/03 Add -lutil to the default FreeBSD LIBS setting.
912 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
913 errors. Otherwise a message that provokes a temporary error (when other
914 messages do not) can cause a whole host to time out.
916 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
917 addresses were routed directly to files or pipes from a redirect router.
918 File deliveries just didn't batch; pipe deliveries might have suffered
921 PH/06 A failure to get a lock for a hints database would erroneously always say
922 "Failed to get write lock", even when it was really a read lock.
924 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
925 of 0600. This has been changed to use the value of the lockfile_mode
926 option (which defaults to 0600).
928 PH/08 Applied small patch from the Sieve maintainer.
930 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
931 folder from quota calculations, a direct delivery into this folder messed
932 up the contents of the maildirsize file. This was because the regex was
933 used only to exclude .Trash (or whatever) when the size of the mailbox
934 was calculated. There was no check that a delivery was happening into an
935 excluded directory. This bug has been fixed by ignoring all quota
936 processing for deliveries into excluded directories.
938 PH/10 Added the maildirfolder_create_regex option to appendfile.
944 PH/01 The code for finding all the local interface addresses on a FreeBSD
945 system running IPv6 was broken. This may well have applied to all BSD
946 systems, as well as to others that have similar system calls. The broken
947 code found IPv4 interfaces correctly, but gave incorrect values for the
948 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
949 that it would not match correctly against @[] and not recognize the IPv6
952 PH/02 The ipliteral router was not recognizing addresses of the form user@
953 [ipv6:....] because it didn't know about the "ipv6:" prefix.
955 PH/03 Added disable_ipv6.
957 PH/04 Changed $reply_address to use the raw form of the headers instead of the
958 decoded form, because it is most often used to construct To: headers
959 lines in autoreplies, and the decoded form may well be syntactically
960 invalid. However, $reply_address has leading white space removed, and all
961 newlines turned into spaces so that the autoreply transport does not
964 PH/05 If group was specified without a user on a router, and no group or user
965 was specified on a transport, the group from the router was ignored.
967 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
968 for visible compile-time settings that can be used to change these
969 numbers, for those that want even more. Backwards compatibility with old
970 spool files has been maintained. However, going back to a previous Exim
971 release will lost any variables that are in spool files.
973 PH/07 Two small changes when running in the test harness: increase delay when
974 passing a TCP/IP connection to a new process, in case the original
975 process has to generate a bounce, and remove special handling of
976 127.0.0.2 (sic), which is no longer necessary.
978 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
979 be the same on different OS.
981 PH/09 Moved a debug statement in filter processing to avoid a race problem when
984 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
985 whether --show-vars was specified or not
987 JJ/02 exipick: Added support for new ACL variable spool format introduced
990 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
991 syntactically invalid From: or Reply-to: line, and a filter used this to
992 generate an autoreply, and therefore failed to obtain an address for the
993 autoreply, Exim could try to deliver to a non-existent relative file
994 name, causing unrelated and misleading errors. What now happens is that
995 it logs this as a hard delivery error, but does not attempt to create a
998 PH/11 The exinext utility has a -C option for testing purposes, but although
999 the given file was scanned by exinext itself; it wasn't being passed on
1000 when Exim was called.
1002 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1003 an end-of-file indication when reading a command response.
1005 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1006 compiled. In many other places in Exim, IPv6 addresses are always
1007 recognized, so I have changed this. It also means that IPv4 domain
1008 literals of the form [IPV4:n.n.n.n] are now always recognized.
1010 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1011 used if the router is not running as root, for example, when verifying at
1012 ACL time, or when using -bh. The debugging output from this situation was
1013 non-existent - all you got was a failure to exec. I have made two
1016 (a) Failures to set uid/gid, the current directory, or a process leader
1017 in a subprocess such as that created by queryprogram now generate
1018 suitable debugging ouput when -d is set.
1020 (b) The queryprogram router detects when it is not running as root,
1021 outputs suitable debugging information if -d is set, and then runs
1022 the subprocess without attempting to change uid/gid.
1024 PH/15 Minor change to Makefile for building test_host (undocumented testing
1027 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1028 additional section of a DNS packet that returns MX or SRV records.
1029 Instead, it always explicitly searches for A/AAAA records. This avoids
1030 major problems that occur when a DNS server includes only records of one
1031 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1032 fixed another bug: if SRV records were looked up and the corresponding
1033 address records were *not* found in the additional section, the port
1034 values from the SRV records were lost.
1036 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1037 using the correct key (the original address) when searching the retry
1038 rules in order to find which one to use for generating the retry hint.
1040 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1041 adding the default one. Similarly, if it contains a Reply-To: header, the
1042 errors_reply_to option, if set, is not used.
1044 PH/19 When calculating a retry time, Exim used to measure the "time since
1045 failure" by looking at the "first failed" field in the retry record. Now
1046 it does not use this if it is later than than the arrival time of the
1047 message. Instead it uses the arrival time. This makes for better
1048 behaviour in cases where some deliveries succeed, thus re-setting the
1049 "first failed" field. An example is a quota failure for a huge message
1050 when small messages continue to be delivered. Without this change, the
1051 "time since failure" will always be short, possible causing more frequent
1052 delivery attempts for the huge message than are intended.
1053 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1055 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1056 $1, $2, $3) because the numerical variables can be reset during some
1057 expansion items (e.g. "match"), thereby losing the authentication data.
1059 PH/21 Make -bV show the size of off_t variables so that the test suite can
1060 decide whether to run tests for quotas > 2G.
1062 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1063 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1064 filecount value is greater than 2G or if a quota value is greater than 2G
1065 on a system where the size of off_t is not greater than 4, a panic error
1068 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1069 never match. The debug and -bh output now contains an explicit error
1070 message indicating a malformed IPv4 address or mask.
1072 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1073 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1074 PH/23 above applies.
1076 PH/25 Do not write to syslog when running in the test harness. The only
1077 occasion when this arises is a failure to open the main or panic logs
1078 (for which there is an explicit test).
1080 PH/26 Added the /no_tell option to "control=freeze".
1082 PH/27 If a host name lookup failed very early in a connection, for example, if
1083 the IP address matched host_lookup and the reverse lookup yielded a name
1084 that did not have a forward lookup, an error message of the form "no IP
1085 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1086 could be logged. Now it outputs the IP address instead of "NULL".
1088 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1089 allows the sender and the authenticated sender to be set when
1090 submitting a message from within Exim. Since child_open_exim() is
1091 documented for local_scan(), the new function should be too.
1093 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1094 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1095 results in an empty string is now treated as unset.
1097 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1099 PH/31 Added sender_verify_fail logging option.
1101 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1102 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1103 tidied the source and removed it altogether.
1105 PH/33 When a queue run was abandoned because the load average was too high, a
1106 log line was always written; now it is written only if the queue_run log
1107 selector is set. In addition, the log line for abandonment now contains
1108 information about the queue run such as the pid. This is always present
1109 in "start" and "stop" lines but was omitted from the "abandon" line.
1111 PH/34 Omit spaces between a header name and the colon in the error message that
1112 is given when verify = headers_syntax fails (if there are lots of them,
1113 the message gets confusing).
1115 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1116 names, as there are now some PTR records that contain slashes. This check
1117 is only to protect against broken name servers that fall over on strange
1118 characters, so the fact that it applies to all lookups doesn't matter.
1120 PH/36 Now that the new test suite is complete, we can remove some of the
1121 special code in Exim that was needed for the old test suite. For example,
1122 sorting DNS records because real resolvers return them in an arbitrary
1123 order. The new test suite's fake resolver always returns records in the
1126 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1127 bounces) except when queue_only is set, to avoid logging races between
1128 the different processes.
1130 PH/38 Panic-die if .include specifies a non-absolute path.
1132 PH/39 A tweak to the "H" retry rule from its user.
1134 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1135 a label. They prevented compilation on older perls.
1137 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1138 a warning to be raised on newish perls.
1140 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1141 on queue. Changes to match documented behaviour of showing count of
1142 messages matching specified criteria.
1144 PH/40 Changed the default ident timeout from 30s to 5s.
1146 PH/41 Added support for the use of login_cap features, on those BSD systems
1147 that have them, for controlling the resources used by pipe deliveries.
1149 PH/42 The content-scanning code uses fopen() to create files in which to put
1150 message data. Previously it was not paying any attention to the mode of
1151 the files. Exim runs with umask(0) because the rest of the code creates
1152 files with open(), and sets the required mode explicitly. Thus, these
1153 files were ending up world-writeable. This was not a big issue, because,
1154 being within the spool directory, they were not world-accessible. I have
1155 created a function called modefopen, which takes an additional mode
1156 argument. It sets umask(777), creates the file, chmods it to the required
1157 mode, then resets the umask. All the relevant calls to fopen() in the
1158 content scanning code have been changed to use this function.
1160 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1161 to 24 hours. This avoids potential overflow problems when processing G
1162 and H retry rules. I suspect nobody ever tinkers with this value.
1164 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1166 PH/45 When the plaintext authenticator is running as a client, the server's
1167 challenges are checked to ensure they are valid base64 strings. By
1168 default, the authentication attempt is cancelled if an invalid string is
1169 received. Setting client_ignore_invalid_base64 true ignores these errors.
1170 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1171 they are received. Thus, the responses can be made to depend on the
1172 challenges. If an invalid string is ignored, an empty string is placed in
1175 PH/46 Messages that are created by the autoreply transport now contains a
1176 References: header, in accordance with RFCs 2822 and 3834.
1178 PH/47 Added authenticated_sender_force to the smtp transport.
1180 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1182 PH/49 Installed latest patch from the Sieve maintainer.
1184 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1185 was also set, the appendfile transport was unnecessarily scanning a
1186 directory of message files (e.g. for maildir delivery) to find the count
1187 of files (along with the size), even though it did not need this
1188 information. It now does the scan only if it needs to find either the
1189 size of the count of files.
1191 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1193 PH/52 Two bugs concerned with error handling when the smtp transport is
1196 (i) Exim was not creating retry information for temporary errors given
1197 for individual recipients after the DATA command when the smtp transport
1198 was used in LMTP mode. This meant that they could be retried too
1199 frequently, and not timed out correctly.
1201 (ii) Exim was setting the flag that allows error details to be returned
1202 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1203 recipients that were returned after the DATA command.
1205 PH/53 This is related to PH/52, but is more general: for any failing address,
1206 when detailed error information was permitted to be returned to the
1207 sender, but the error was temporary, then after the final timeout, only
1208 "retry timeout exceeded" was returned. Now it returns the full error as
1209 well as "retry timeout exceeded".
1211 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1212 do this, and (what is worse) MTAs that accept it.
1214 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1215 will now be deprecated.
1217 PH/56 New os.c-cygwin from the Cygwin maintainer.
1219 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1220 formats (previously only available in exim formats via -bpr, -bpru,
1221 and -bpra. Now also available in native and exiqgrep formats)
1223 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1224 with very large, slow to parse queues
1226 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1228 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1230 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1231 responses to authentication challenges, though it was showing the
1232 challenges; (ii) I've removed the CR characters from the debug output for
1235 PH/58 Allow for the insertion of a newline as well as a space when a string
1236 is turned into more than one encoded-word during RFC 2047 encoding. The
1237 Sieve code now uses this.
1239 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1240 data_4xx, lost_connection, tls_required.
1242 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1243 message was being sent as an SMTP response.
1245 PH/61 Add -l and -k options to exicyclog.
1247 PH/62 When verifying, if an address was redirected to one new address, so that
1248 verification continued, and the new address failed or deferred after
1249 having set something in $address_data, the value of $address_data was not
1250 passed back to the ACL. This was different to the case when no
1251 redirection occurred. The value is now passed back in both cases.
1253 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1254 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1255 use login_cap.h, so on its own it isn't the distinguishing feature. The
1256 new name refers directly to the setclassresources() function.
1258 PH/65 Added configuration files for NetBSD3.
1260 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1262 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1263 is preferred over IPv4.
1265 PH/68 The bounce_return_message and bounce_return_body options were not being
1266 honoured for bounces generated during the reception of non-SMTP messages.
1267 In particular, this applied to messages rejected by the ACL. This bug has
1268 been fixed. However, if bounce_return_message is true and bounce_return_
1269 body is false, the headers that are returned for a non-SMTP message
1270 include only those that have been read before the error was detected.
1271 (In the case of an ACL rejection, they have all been read.)
1273 PH/69 The HTML version of the specification is now built in a directory called
1274 spec_html instead of spec.html, because the latter looks like a path with
1275 a MIME-type, and this confuses some software.
1277 PH/70 Catch two compiler warnings in sieve.c.
1279 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1280 function verify_get_ident() calls ip_connect() to connect a socket, but
1281 if the "connect()" function timed out, ip_connect() used to close the
1282 socket. However, verify_get_ident() also closes the socket later, and in
1283 between Exim writes to the log, which may get opened at this point. When
1284 the socket was closed in ip_connect(), the log could get the same file
1285 descriptor number as the socket. This naturally causes chaos. The fix is
1286 not to close the socket in ip_connect(); the socket should be closed by
1287 the function that creates it. There was only one place in the code where
1288 this was missing, in the iplookup router, which I don't think anybody now
1289 uses, but I've fixed it anyway.
1291 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1292 well as to direct DNS lookups. Otherwise the handling of names in host
1293 lists is inconsistent and therefore confusing.
1299 PH/01 Two changes to the default runtime configuration:
1301 (1) Move the checks for relay_from_hosts and authenticated clients from
1302 after to before the (commented out) DNS black list checks.
1304 (2) Add control=submission to the relay_from_hosts and authenticated
1305 clients checks, on the grounds that messages accepted by these
1306 statements are most likely to be submissions.
1308 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1310 (1) Generate an error if the third argument for the ${prvs expansion is
1313 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1316 (3) Reset the variables that are obtained from the first argument of
1317 ${prvscheck and used in the second argument before leaving the code,
1318 because their memory is reclaimed, so using them afterwards may do
1321 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1322 one (it's much easier than Tom thought :-).
1324 (5) Because of (4), we can now allow for the use of $prvscheck_result
1325 inside the third argument.
1327 PH/03 For some reason, the default setting of PATH when running a command from
1328 a pipe transport was just "/usr/bin". I have changed it to
1331 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1332 anything to be listed in the output from -bV.
1334 PH/05 When a filter generated an autoreply, the entire To: header line was
1335 quoted in the delivery log line, like this:
1337 => >A.N.Other <ano@some.domain> <original@ddress> ...
1339 This has been changed so that it extracts the operative address. There
1340 may be more than one such address. If so, they are comma-separated, like
1343 => >ano@some.domain,ona@other.domain <original@ddress> ...
1345 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1346 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1347 not being looked up in the rDNS to get a host name, Exim was showing the
1348 IP address twice in Received: lines, even though the IP addresses were
1349 identical. For example:
1351 Received: from [1.2.3.4] (helo=[1.2.3.4])
1353 However, if the real host name was known, it was omitting the HELO data
1354 if it matched the actual IP address. This has been tidied up so that it
1355 doesn't show the same IP address twice.
1357 PH/07 When both +timestamp and +memory debugging was on, the value given by
1358 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1359 called by the debug printing, thereby overwriting the timestamp buffer.
1360 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1362 PH/08 When the original message was included in an autoreply transport, it
1363 always said "this is a copy of the message, including all the headers",
1364 even if body_only or headers_only was set. It now gives an appropriate
1367 PH/09 Applied a patch from the Sieve maintainer which:
1369 o fixes some comments
1370 o adds the (disabled) notify extension core
1371 o adds some debug output for the result of if/elsif tests
1372 o points to the current vacation draft in the documentation
1373 and documents the missing references header update
1377 o fixes a bug in processing the envelope test (when testing
1378 multiple envelope elements, the last element determinted the
1381 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1382 Electronic Mail") by including:
1384 Auto-submitted: auto-generated
1386 in the messages that it generates (bounce messages and others, such as
1387 warnings). In the case of bounce messages for non-SMTP mesages, there was
1388 also a typo: it was using "Auto_submitted" (underscore instead of
1389 hyphen). Since every message generated by Exim is necessarily in response
1390 to another message, thes have all been changed to:
1392 Auto-Submitted: auto-replied
1394 in accordance with these statements in the RFC:
1396 The auto-replied keyword:
1398 - SHOULD be used on messages sent in direct response to another
1399 message by an automatic process,
1401 - MUST NOT be used on manually-generated messages,
1403 - MAY be used on Delivery Status Notifications (DSNs) and Message
1404 Disposition Notifications (MDNs),
1406 - MUST NOT be used on messages generated by automatic or periodic
1407 processes, except for messages which are automatic responses to
1410 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1411 to the default Received: header definition.
1413 PH/12 Added log selector acl_warn_skipped (default on).
1415 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1416 variables because (a) they are in the wrong storage pool and (b) even if
1417 they were copied, it wouldn't work properly because of the caching.
1419 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1420 checking when decoding. Apparently there are clients that generate
1421 overlong encoded strings. Why am I not surprised?
1423 PH/15 If the first argument of "${if match_address" was not empty, but did not
1424 contain an "@" character, Exim crashed. Now it writes a panic log message
1425 and treats the condition as false.
1427 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1429 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1430 extension "envelope-auth". The code is finished and in agreement with
1431 other implementations, but there is no documentation so far and in fact,
1432 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1433 not changing the active code.
1435 Print executed "if" and "elsif" statements when debugging is used. This
1436 helps a great deal to understand what a filter does.
1438 Document more things not specified clearly in RFC3028. I had all this
1439 sorted out, when out of a sudden new issues came to my mind. Oops."
1441 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1444 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1445 canonical form (as documented). However, after a host name lookup from
1446 the IP address, check_host() was doing a simple string comparison with
1447 addresses acquired from the DNS when checking that the found name did
1448 have the original IP as one of its addresses. Since any found IPv6
1449 addresses are likely to be in abbreviated form, the comparison could
1450 fail. Luckily, there already exists a function for doing the comparison
1451 by converting both addresses to binary, so now that is used instead of
1452 the text comparison.
1454 PH/20 There was another similar case to PH/19, when a complete host name was
1455 given in a host list; looking up its IP address could give an abbreviated
1456 form, whereas the current host's name might or might not be abbreviated.
1457 The same fix has been applied.
1463 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1464 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1467 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1468 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1470 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1472 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1473 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1474 Also, for "Auto-Submitted", treat anything other than "no" as
1475 non-personal, in accordance with RFC 3834. (Previously it treated
1476 anything starting "auto-" as non-personal.)
1478 TF/01 The control=submission/name=... option had a problem with syntax
1479 errors if the name included a slash character. The /name= option
1480 now slurps the rest of the string, so it can include any characters
1481 but it must come last in the list of options (after /sender_retain
1484 PH/05 Some modifications to the interface to the fake nameserver for the new
1492 TK/01 Added the "success_on_redirect" address verification option. See
1493 NewStuff for rationale and an example.
1495 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1497 PH/02 Patch to exigrep to allow it to work on syslog lines.
1499 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1500 fread() to skip over the body file's header line, because in Cygwin the
1501 header line is locked and is inaccessible.
1503 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1504 co-exist for some time) to make it clear that it is the Exim ID that is
1505 referenced, not the Message-ID: header line.
1507 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1508 string_format() function, because snprintf() does not exist on all
1511 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1512 ${stat: expansion item.
1514 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1515 protocol synchronization error", to keep the pedants happy.
1517 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1518 well as for IRIX systems, when gcc is being used. See the host.c source
1521 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1523 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1526 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1527 transport and to the smtp transport in LMTP mode.
1529 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1531 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1532 run a filter in a subprocess. This could lead to confusion in subsequent
1533 lookups in the parent process. There should also be a search_tidyup() at
1534 the end of the subprocess.
1536 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1537 only if the host matched helo_try_verify_hosts, which caused the
1538 verification to occur when the EHLO/HELO command was issued. The ACL just
1539 tested the remembered result. Now, if a previous verification attempt has
1540 not happened, "verify = helo" does it there and then.
1542 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1544 TK/03 Fix log output including CR from clamd.
1546 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1547 exist provoked a memory error which could cause a segfault.
1549 PH/15 Installed PCRE 6.2
1551 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1553 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1554 of the problem. Specifically, suggested +O2 rather than +O1 for the
1557 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1559 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1560 record, the port was not correctly logged when the outgoing_port log
1561 selector was set (it logged the transort's default port).
1563 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1564 fallback_hosts, and "hosts" in the smtp transport.
1566 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1567 host errors such as "Connection refused".
1569 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1570 authentication with radiusclient 0.4.9:
1572 - Error returned from rc_read_config was caught wrongly
1573 - Username/password not passed on to radius server due to wrong length.
1575 The presumption is that some radiusclient API changes for 4.51/PH/17
1576 were not taken care of correctly. The code is still untested by me (my
1577 Linux distribution still has 0.3.2 of radiusclient), but it was
1578 contributed by a Radius user.
1580 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1581 expanding the "port" option of the smtp transport.
1583 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1584 while reading a MIME header. Thanks to Tom Hughes for a patch.
1586 PH/24 Include config.h inside local_scan.h so that configuration settings are
1589 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1590 that in an ACL for RCPT (for example), you can examine exactly what was
1593 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1594 commands, but it was not correctly comparing the address with the actual
1595 client host address. Thus, it would show the EHLO address in Received:
1596 header lines when this was not necessary.
1598 PH/27 Added the % operator to ${eval:}.
1600 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1601 it should be ignoring failures (because with -C, for example, it has lost
1602 privilege). It wasn't ignoring creation failures other than "already
1605 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1608 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1609 because an input line was too long, either on its own, or by virtue of
1610 too many continuations, the temporary file was not being removed, and the
1611 return code was incorrect.
1613 PH/31 Missing "BOOL" in function definition in filtertest.c.
1615 PH/32 Applied Sieve patches from the maintainer.
1617 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1619 PH/33 Added "verify = not_blind".
1621 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1622 Local/Makefile (with some defaults set). These are used in built scripts
1623 such as exicyclog, but they have never been used in the exim_install
1624 script (though there are many overriding facilities there). I have
1625 arranged that the exim_install script now takes note of these two
1628 PH/35 Installed configuration files for Dragonfly.
1630 PH/36 When a locally submitted message by a trusted user did not contain a
1631 From: header, and the sender address was obtained from -f or from an SMTP
1632 MAIL command, and the trusted user did not use -F to supply a sender
1633 name, $originator_name was incorrectly used when constructing a From:
1634 header. Furthermore, $originator_name was used for submission mode
1635 messages from external hosts without From: headers in a similar way,
1636 which is clearly wrong.
1638 PH/37 Added control=suppress_local_fixups.
1640 PH/38 When log_selector = +received_sender was set, and the addition of the
1641 sender made the log line's construction buffer exactly full, or one byte
1642 less than full, an overflow happened when the terminating "\n" was
1645 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1646 when the result of a list match is failure because a DNS lookup failed.
1648 PH/40 RM_COMMAND is now used in the building process.
1650 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1651 the "build-* directories that it finds.
1653 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1654 address in a domain literal was a prefix of an interface address.
1656 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1657 when verifying a sender address, unless rewrite_headers is false.
1659 PH/44 Wrote a long comment about why errors_to addresses are verified as
1660 recipients, not senders.
1662 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1663 the ratelimit ACL was added.
1665 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1667 PH/46 Added extra information about PostgreSQL errors to the error string.
1669 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1670 avoiding the need to install special zones in a real server. This is
1671 backwards compatible; if it can't find the fake resolver, it drops back.
1672 Thus, both old and new test suites can be run.
1674 TF/02 Added util/ratelimit.pl
1676 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1677 clock is set back in time.
1679 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1680 Candler <B.Candler@pobox.com>.
1682 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1683 OK for addresses that are the result of redirections.
1685 PH/48 A number of further additions for the benefit of the new test suite,
1686 including a fake gethostbyname() that interfaces to the fake DNS resolver
1689 TF/06 The fix for widen_domains has also been applied to qualify_single and
1690 search_parents which are the other dnslookup options that can cause
1693 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1696 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1698 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1699 service" if it encountered an MX record with an empty target hostname.
1700 The message is now "an MX or SRV record indicated no SMTP service".
1702 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1703 if the DNS of the sending site is misconfigured. This is quite a
1704 common situation. This change restores the behaviour of treating a
1705 helo verification defer as a failure.
1707 PH/51 If self=fail was set on a router, the bounce message did not include the
1708 actual error message.
1714 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1716 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1717 command itself ended in error, the underlying message about the transport
1718 filter timeout was being overwritten with the pipe command error. Now the
1719 underlying error message should be appended to the second error message.
1721 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1723 PH/02 Reduce the amount of output that "make" produces by default. Full output
1724 can still be requested.
1726 PH/03 The warning log line about a condition test deferring for a "warn" verb
1727 was being output only once per connection, rather than after each
1728 occurrence (because it was using the same function as for successful
1729 "warn" verbs). This seems wrong, so I have changed it.
1731 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1732 it should not have, which might have caused a crash in the right
1733 circumstances, but probably never did.
1735 PH/04 Installed a modified version of Tony Finch's patch to make submission
1736 mode fix the return path as well as the Sender: header line, and to
1737 add a /name= option so that you can make the user's friendly name appear
1740 TF/03 Added the control = fakedefer ACL modifier.
1742 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1743 Mark Lowes for thorough testing.
1745 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1747 TK/03 Merged latest SRS patch from Miles Wilton.
1749 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1750 with the definition in sysexits.h (which is #included earlier).
1751 Fortunately, Exim does not actually use EX_OK. The code used to try to
1752 preserve the sysexits.h value, by assumimg that macro definitions were
1753 scanned for macro replacements. I have been disabused of this notion,
1754 so now the code just undefines EX_OK before #including unistd.h.
1756 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1757 in the smtp transport. When a block could not be written in a single
1758 write() function, the timeout was being re-applied to each part-write.
1759 This seems wrong - if the receiver was accepting one byte at a time it
1760 would take for ever. The timeout is now adjusted when this happens. It
1761 doesn't have to be particularly precise.
1763 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1764 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1766 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1767 without a domain if the check to <postmaster@domain> fails.
1769 SC/01 Eximstats: added -xls and the ability to specify output files
1770 (patch written by Frank Heydlauf).
1772 SC/02 Eximstats: use FileHandles for outputing results.
1774 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1776 SC/04 Eximstats: fixed display of large numbers with -nvr option
1778 SC/05 Eximstats: fixed merging of reports with empty tables.
1780 SC/06 Eximstats: added the -include_original_destination flag
1782 SC/07 Eximstats: removed tabs and trailing whitespace.
1784 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1786 TK/06 MBOX spool code: Add real "From " MBOX separator line
1787 so the .eml file is really in mbox format (even though
1788 most programs do not really care). Patch from Alex Miller.
1790 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1791 The latter is generated from $received_to and is only set if the
1792 message has one envelope recipient. SA can use these headers,
1793 obviously out-of-the-box. Patch from Alex Miller.
1795 PH/08 The ${def test on a variable was returning false if the variable's
1796 value was "0", contrary to what the specification has always said!
1797 The result should be true unless the variable is empty.
1799 PH/09 The syntax error of a character other than { following "${if
1800 def:variable_name" (after optional whitespace) was not being diagnosed.
1801 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1802 accidental colon was present, for example, could give incorrect results.
1804 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1805 result is used (not including appendfile, where other changes are about
1808 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1809 This involved changing a lot of size variables from int to off_t. It
1810 should work with maildirs and everything.
1812 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1813 spamd dying while we are connected to it.
1815 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1818 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1819 function for BDB 4.3.
1821 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1823 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1824 the sources for PCRE have been re-arranged and more files are now
1827 PH/15 The code I had for printing potentially long long variables in PH/11
1828 above was not the best (it lost precision). The length of off_t variables
1829 is now inspected at build time, and an appropriate printing format (%ld
1830 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1831 to be "long long int" or "long int". This is needed for the internal
1832 formatting function string_vformat().
1834 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1835 the configuration file to be ":syslog", then the script "guesses" where
1836 the logs files are, rather than using the compiled in default. In our
1837 case the guess is not the same as the compiled default, so the script
1838 suddenly stopped working when I started to use syslog. The patch checks
1839 to see if log_file_path is "". If so, it attempts to read it from exim
1840 with no configuration file to get the compiled in version, before it
1841 falls back to the previous guessing code."
1843 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1844 implementing BATV in an Exim configuration. See NewStuff for the gory
1847 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1848 Makefile that are specific to HP-UX.
1850 PH/18 If the "use_postmaster" option was set for a recipient callout together
1851 with the "random" option, the postmaster address was used as the MAIL
1852 FROM address for the random test, but not for the subsequent recipient
1853 test. It is now used for both.
1855 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1856 patch removes a few documentation additions to RFC 3028, because the
1857 latest draft now contains them. It adds the new en;ascii-case comparator
1858 and a new error check for 8bit text in MIME parts. Comparator and
1859 require names are now matched exactly. I enabled the subaddress
1860 extension, but it is not well tested yet (read: it works for me)."
1862 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1863 rework some of the code of TK/09 above to avoid the hardwired use of
1864 "%lld" and "long long". Replaced the call to snprintf() with a call to
1867 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1868 records point to non-existent hosts", "retry timeout exceeded", and
1869 "retry time not reached for any host after a long failure period".
1871 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1872 experimental DomainKeys support:
1874 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1875 (2) On an error such as an illegally used "control", the wrong name for
1876 the control was given.
1878 These problems did NOT occur unless DomainKeys support was compiled.
1880 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1882 PH/24 Added ${if match_ip condition.
1884 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1885 databases so that it will be absolutely obvious if a crash occurs in the
1886 DB library. This is a regular occurrence (often caused by mis-matched
1889 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1890 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1891 picked up on a user's system that detects such things. There doesn't seem
1892 to be a gcc warning option for this - only an attribute that has to be
1893 put on the function's prototype. It seems that in Fedora Core 4 they have
1894 set this on a number of new functions. No doubt there will be more in due
1897 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1898 specify a transport. However, if an address that was verified by such a
1899 router was the subject of a callout, Exim crashed because it tried to
1900 read the rcpt_include_affixes from the non-existent transport. Now it
1901 just assumes that the setting of that option is false. This bug was
1902 introduced by 4.51/PH/31.
1904 PH/28 Changed -d+all to exclude +memory, because that information is very
1905 rarely of interest, but it makes the output a lot bigger. People tend to
1906 do -d+all out of habit.
1908 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1909 code in os-type was giving problems when libc.so lives in lib64, like on
1912 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1913 aren't the modern standard, and it seems that some systems' include files
1914 don't always have them. Exim was already checking for some of the newer
1915 ones like T_AAAA, and defining it itself. I've added checks for all the
1916 record types that Exim uses.
1918 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1919 not automatically generating a new one, as it is supposed to. This
1920 prevented TLS from working. If the file did exist, but contained invalid
1921 data, a new version was generated, as expected. It was only the case of a
1922 non-existent file that was broken.
1924 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1925 with a change in libdomainkeys > 0.64.
1927 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1928 from DNS. If the selector record carries the flag, it now has
1929 precedence over the domain-wide flag.
1931 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1933 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1934 the use of a port name that isn't defined in /etc/services) occurred, the
1935 message was deferred as in a normal delivery, and thus remained on the
1936 spool, instead of being failed because of the mua_wrapper setting. This
1937 is now fixed, and I tidied up some of the mua_wrapper messages at the
1940 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1941 the messages in a hash of arrays rather than using individual hashes.
1942 This is a bit cleaner and results in dramatic memory savings, albeit
1943 at a slight CPU cost.
1945 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1946 as requested by Marc Sherman.
1948 SC/10 Eximstats: added histograms for user specified patterns as requested
1951 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1953 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1954 fopen() in the content-scanning modules that did not already have it.
1960 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1961 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1963 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1965 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1967 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1968 to test Sieve filters that use "vacation".
1970 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1971 that changes the way the GnuTLS parameters are stored in the cache file.
1972 The new format can be generated externally. For backward compatibility,
1973 if the data in the cache doesn't make sense, Exim assumes it has read an
1974 old-format file, and it generates new data and writes a new file. This
1975 means that you can't go back to an older release without removing the
1978 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1979 work if there are any delivery delays because "one_time" forces the
1980 parent to be marked "delivered", so its unseen clone is never tried
1981 again. For this reason, Exim now forbids the simultaneous setting of
1984 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1985 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1986 entry if you want to know the details. The fix, however, neglected to
1987 consider the case when local delivery batching is involved. The test for
1988 "previously delivered" was not happening when checking to see if an
1989 address could be batched with a previous (undelivered) one; under
1990 certain circumstances this could lead to multiple deliveries to the same
1993 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1994 in its include files, and this causes problems building Exim.
1996 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1997 header_syntax) but Exim was just ignoring anything given after a slash.
1998 In particular, this caused confusion with an attempt to use "verify =
1999 reverse_host_lookup/defer_ok". An error is now given when options are
2000 supplied for verify items that do not have them. (Maybe reverse_host_
2001 lookup should have a defer_ok option, but that's a different point.)
2003 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2004 defined by RFC 821) to 2048, because there were problems with some AUTH
2005 commands, and RFC 1869 says the size should be increased for extended
2006 SMTP commands that take arguments.
2008 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2011 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2012 "unknown" error; now it says that the functionality isn't in the binary.
2014 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2015 an address' error message when a string expansion fails (syntax or
2016 whatever). Otherwise the password may appear in the log. Following change
2017 PH/42 below, there is no longer a chance of it appearing in a bounce
2020 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2022 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2023 of its addresses was ever tried. (Bugzilla bug #2.)
2025 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2026 the result incorrectly in the debug output. (It correctly added a newline
2027 to what was transported.)
2029 TF/01 Added $received_time.
2031 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2032 commented out examples of how to interface to a virus scanner and to
2033 SpamAssassin. Also added commented examples of av_scanner and
2034 spamd_address settings.
2036 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2037 and controls are allowed in which ACLs. There were a couple of minor
2038 errors. Some of the entries in the conditions table (which is a table of
2039 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2040 negation of where the condition IS allowed.
2042 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2044 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2045 header file does not have a version number, so I've had to invent a new
2046 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2047 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2048 radiusclient), but it was contributed by a Radius user.
2050 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2051 files or directories, for OpenSSL.
2053 PH/19 When an Exim process that is running as root has to create an Exim log
2054 file, it does so in a subprocess that runs as exim:exim so as to get the
2055 ownership right at creation (otherwise, other Exim processes might see
2056 the file with the wrong ownership). There was no test for failure of this
2057 fork() call, which would lead to the process getting stuck as it waited
2058 for a non-existent subprocess. Forks do occasionally fail when resources
2059 run out. I reviewed all the other calls to fork(); they all seem to check
2062 PH/20 When checking for unexpected SMTP input at connect time (before writing
2063 the banner), Exim was not dealing correctly with a non-positive return
2064 from the read() function. If the client had disconnected by this time,
2065 the result was a log entry for a synchronization error with an empty
2066 string after "input=" when read() returned zero. If read() returned -1
2067 (an event I could not check), uninitialized data bytes were printed.
2068 There were reports of junk text (parts of files, etc) appearing after
2071 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2073 PH/22 Added support for macro redefinition, and (re)definition in between
2074 driver and ACL definitions.
2076 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2077 forgetting to use the resulting value; it was using the unexpanded value.
2079 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2080 hadn't been configured. The fix is from Juergen Kreileder, who
2081 understands it better than I do:
2083 "Here's what I see happening with three configured cyrus_sasl
2084 authenticators configured (plain, login, cram-md5):
2086 On startup auth_cyrus_sasl_init() gets called for each of these.
2087 This means three calls to sasl_listmech() without a specified mech_list.
2088 => SASL tests which mechs of all available mechs actually work
2089 => three warnings about OTP not working
2090 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2092 With the patch, sasl_listmech() also gets called three times. But now
2093 SASL's mech_list option is set to the server_mech specified in the the
2094 authenticator. Or in other words, the answer from sasl_listmech()
2095 gets limited to just the mech you're testing for (which is different
2097 => the return list contains just 'plain' or 'login', 'cram-md5' or
2098 nothing depending on the value of ob->server_mech.
2100 I've just tested the patch: Authentication still works fine,
2101 unavailable mechs specified in the exim configuration are still
2102 caught, and the auth.log warnings about OTP are gone."
2104 PH/25 When debugging is enabled, the contents of the command line are added
2105 to the debugging output, even when log_selector=+arguments is not
2108 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2109 answer is "GNU", and only if the return is "GNU/something" is the answer
2112 PH/27 $acl_verify_message is now set immediately after the failure of a
2113 verification in an ACL, and so is available in subsequent modifiers. In
2114 particular, the message can be preserved by coding like this:
2116 warn !verify = sender
2117 set acl_m0 = $acl_verify_message
2119 Previously, $acl_verify_message was set only while expanding "message"
2120 and "log_message" when a very denied access.
2122 PH/28 Modified OS/os.c-Linux with
2124 -#ifndef OS_LOAD_AVERAGE
2125 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2127 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2128 nomenclature these days.)
2130 PH/29 Installed patch from the Sieve maintainer that adds the options
2131 sieve_useraddress and sieve_subaddress to the redirect router.
2133 PH/30 In these circumstances:
2134 . Two addresses routed to the same list of hosts;
2135 . First host does not offer TLS;
2136 . First host accepts first address;
2137 . First host gives temporary error to second address;
2138 . Second host offers TLS and a TLS session is established;
2139 . Second host accepts second address.
2140 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2141 and peerdn, if requested) that were in fact used only for the second
2144 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2145 attention to any local part prefix or suffix that was matched by the
2146 router that accepted the address. It now behaves in the same way as it
2147 does for delivery: the affixes are removed from the local part unless
2148 rcpt_include_affixes is set on the transport.
2150 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2151 timeout during the DATA phase of an incoming message.
2153 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2154 applied a patch sanctioned by the Sieve maintainer.
2156 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2157 the uid or gid is negative. A case of a negative gid caused this to be
2158 noticed. The fix allows for either to be negative.
2160 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2161 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2164 PH/36 The change PH/12 above was broken. Fixed it.
2166 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2167 the grounds that routing the same address twice would always produce the
2168 same answer. This might have been true once, but it is certainly no
2169 longer true now. Routing a child address may depend on the previous
2170 routing that produced that child. Some complicated redirection strategies
2171 went wrong when messages had multiple recipients, and made Exim's
2172 behaviour dependent on the order in which the addresses were given.
2174 I have moved the duplicate checking until after the routing is complete.
2175 Exim scans the addresses that are assigned to local and remote
2176 transports, and removes any duplicates. This means that more work will be
2177 done, as duplicates will always all be routed, but duplicates are
2178 presumably rare, so I don't expect this is of any significance.
2180 For deliveries to pipes, files, and autoreplies, the duplicate checking
2181 still happens during the routing process, since they are not going to be
2184 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2185 It corrects a timeout issue with spamd. This is Ian's comment: "The
2186 background is that sometimes spamd either never reads data from a
2187 connection it has accepted, or it never writes response data. The exiscan
2188 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2189 blindly assumes that writes won't block so it may never time out."
2191 PH/39 Allow G after quota size as well as K and M.
2193 PH/40 The value set for $authenticated_id in an authenticator may not contain
2194 binary zeroes or newlines because the value is written to log lines and
2195 to spool files. There was no check on this. Now the value is run through
2196 the string_printing() function so that such characters are converted to
2197 printable escape sequences.
2199 PH/41 $message_linecount is a new variable that contains the total number of
2200 lines in the message. Compare $body_linecount, which is the count for the
2203 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2204 bounce and delay warning messages, except in certain special cases, which
2207 (a) An SMTP error message from a remote host;
2208 (b) A message specified in a :fail: redirection;
2209 (c) A message specified in a "fail" command in a system filter;
2210 (d) A message specified in a FAIL return from the queryprogram router;
2211 (e) A message specified by the cannot_route_message router option.
2213 In these cases only, Exim does include the error details in bounce and
2214 warning messages. There are also a few cases where bland messages such
2215 as "unrouteable address" or "local delivery error" are given.
2217 PH/43 $value is now also set for the "else" part of a ${run expansion.
2219 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2220 being worked on, but at least Exim now implements the latest version to
2223 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2224 process to complete was treated as a delivery failure, a timeout while
2225 writing the message to the pipe was logged, but erroneously treated as a
2226 successful delivery. Such timeouts include transport filter timeouts. For
2227 consistency with the overall process timeout, these timeouts are now
2228 treated as errors, giving rise to delivery failures by default. However,
2229 there is now a new Boolean option for the pipe transport called
2230 timeout_defer, which, if set TRUE, converts the failures into defers for
2231 both kinds of timeout. A transport filter timeout is now identified in
2234 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2235 systems where "make" and "gmake" are different, calling "gmake" at top
2236 level broke things. I've arranged for the value of $(MAKE) to be passed
2237 from the Makefile to this script so that it can call the same version of
2241 A note about Exim versions 4.44 and 4.50
2242 ----------------------------------------
2244 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2245 changes of various kinds. As a consequence, a big documentation update was
2246 needed. This delayed the release for rather longer than seemed good, especially
2247 in the light of a couple of (minor) security issues. Therefore, the changes
2248 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2249 release. So 4.44 and 4.50 are in effect two different branches that both start
2252 I have left the 4.50 change log unchanged; it contains all the changes since
2253 4.43. The change log for 4.44 is below; many of its items are identical to
2254 those for 4.50. This seems to be the most sensible way to preserve the
2255 historical information.
2261 1. Minor wording change to the doc/README.SIEVE file.
2263 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2264 computation of the current number of files was incorrect.
2266 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2267 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2270 4. Give more explanation in the error message when the command for a transport
2271 filter fails to execute.
2273 5. There are several places where Exim runs a non-Exim command in a
2274 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2275 was being done only for the command run by the queryprogram router. It is
2276 now done for all such subprocesses. The other cases are: ${run, transport
2277 filters, and the commands run by the lmtp and pipe transports.
2279 6. Added CONFIGURE_GROUP build-time option.
2281 7. Some older OS have a limit of 256 on the maximum number of file
2282 descriptors. Exim was using setrlimit() to set 1000 as a large value
2283 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2284 systems. I've change it so that if it can't get 1000, it tries for 256.
2286 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2287 was an oversight, and furthermore, ever since the addition of extra
2288 controls (e.g. 4.43/32), the checks on when to allow different forms of
2289 "control" were broken. There should now be diagnostics for all cases when a
2290 control that does not make sense is encountered.
2292 9. Added the /retain_sender option to "control=submission".
2294 10. $recipients is now available in the predata ACL (oversight).
2296 11. Tidy the search cache before the fork to do a delivery from a message
2297 received from the command line. Otherwise the child will trigger a lookup
2298 failure and thereby defer the delivery if it tries to use (for example) a
2299 cached ldap connection that the parent has called unbind on.
2301 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2302 of $address_data from the recipient verification was clobbered by the
2303 sender verification.
2305 13. The value of address_data from a sender verification is now available in
2306 $sender_address_data in subsequent conditions in the ACL statement.
2308 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2310 15. Added a new option "connect=<time>" to callout options, to set a different
2313 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2314 was its contents. (It was OK if the option was not defined at all.)
2316 17. A "Completed" log line is now written for messages that are removed from
2317 the spool by the -Mrm option.
2319 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2320 information about exactly what failed.
2322 19. Added -dd to debug only the daemon process.
2324 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2325 handles timeouts, both on the server side and network timeouts. Renamed the
2326 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2328 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2329 It is now set to "smtps".
2331 22. $host_address is now set to the target address during the checking of
2332 ignore_target_hosts.
2334 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2335 being passed; this would have caused $sender_host_name to have been used if
2336 matching the list had actually called for a host name (not very likely,
2337 since this list is usually IP addresses). A host name is now passed as
2340 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2341 code that specifies a non-restarting handler (typically sigaction() in
2342 modern systems) in an attempt to fix a rare and obscure crash bug.
2344 25. Narrowed the window for a race in the daemon that could cause it to ignore
2345 SIGCHLD signals. This is not a major problem, because they are used only to
2346 wake it up if nothing else does.
2348 26. A malformed maildirsize file could cause Exim to calculate negative values
2349 for the mailbox size or file count. Odd effects could occur as a result.
2350 The maildirsize information is now recalculated if the size or filecount
2353 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2354 support for a long time. Removed HAVE_SYS_VFS_H.
2356 28. Installed the latest version of exipick from John Jetmore.
2358 29. In an address list, if the pattern was not a regular expression, an empty
2359 subject address (from a bounce message) matched only if the pattern was an
2360 empty string. Non-empty patterns were not even tested. This was the wrong
2361 because it is perfectly reasonable to use an empty address as part of a
2362 database query. An empty address is now tested by patterns that are
2363 lookups. However, all the other forms of pattern expect the subject to
2364 contain a local part and a domain, and therefore, for them, an empty
2365 address still always fails if the pattern is not itself empty.
2367 30. Exim went into a mad DNS loop when attempting to do a callout where the
2368 host was specified on an smtp transport, and looking it up yielded more
2369 than one IP address.
2371 31. Re-factored the code for checking spool and log partition space into a
2372 function that finds that data and another that does the check. The former
2373 is then used to implement four new variables: $spool_space, $log_space,
2374 $spool_inodes, and $log_inodes.
2376 32. The RFC2047 encoding function was originally intended for short strings
2377 such as real names; it was not keeping to the 75-character limit for
2378 encoded words that the RFC imposes. It now respects the limit, and
2379 generates multiple encoded words if necessary. To be on the safe side, I
2380 have increased the buffer size for the ${rfc2047: expansion operator from
2383 33. It is now permitted to omit both strings after an "if" condition; if the
2384 condition is true, the result is "true". As before, when the second string
2385 is omitted, a false condition yields an empty string. This makes it less
2386 cumbersome to write custom ACL and router conditions.
2388 34. Failure to deliver a bounce message always caused it to be frozen, even if
2389 there was an errors_to setting on the router. The errors_to setting is now
2392 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2393 canonical form (fully expanded) before being placed in
2394 $sender_host_address.
2396 36. The table in the code that translates DNS record types into text (T_A to
2397 "A" for instance) was missing entries for NS and CNAME. It is just possible
2398 that this could have caused confusion if both these types were looked up
2399 for the same domain, because the text type is used as part of Exim's
2400 per-process caching. But the chance of anyone hitting this buglet seems
2403 37. The dnsdb lookup has been extended in a number of ways.
2405 (1) There is a new type, "zns", which walks up the domain tree until it
2406 finds some nameserver records. It should be used with care.
2408 (2) There is a new type, "mxh", which is like "mx" except that it returns
2409 just the host names, not the priorities.
2411 (3) It is now possible to give a list of domains (or IP addresses) to be
2412 looked up. The behaviour when one of the lookups defers can be
2413 controlled by a keyword.
2415 (4) It is now possible to specify the separator character for use when
2416 multiple records are returned.
2418 38. The dnslists ACL condition has been extended: it is now possible to supply
2419 a list of IP addresses and/or domains to be looked up in a particular DNS
2422 39. Added log_selector=+queue_time_overall.
2424 40. When running the queue in the test harness, wait just a tad after forking a
2425 delivery process, to get repeatability of debugging output.
2427 41. Include certificate and key file names in error message when GnuTLS fails
2428 to set them up, because the GnuTLS error message doesn't include the name
2429 of the failing file when there is a problem reading it.
2431 42. Allow both -bf and -bF in the same test run.
2433 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2435 44. The "Exiscan patch" is now merged into the mainline Exim source.
2437 45. Sometimes the final signoff response after QUIT could fail to get
2438 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2439 < 0 before doing a fflush(). This bug looks as though it goes back to the
2440 introduction of TLS in release 3.20, but "sometimes" must have been rare
2441 because the tests only now provoked it.
2443 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2444 (this can affect the format of dates).
2446 47. exim_tidydb, when checking for the continued existence of a message for
2447 which it has found a message-specific retry record, was not finding
2448 messages that were in split spool directories. Consequently, it was
2449 deleting retry records that should have stayed in existence.
2451 48. Steve fixed some bugs in eximstats.
2453 49. The SPA authentication driver was not abandoning authentication and moving
2454 on to the next authenticator when an expansion was forced to fail,
2455 contradicting the general specification for all authenticators. Instead it
2456 was generating a temporary error. It now behaves as specified.
2458 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2459 (the order specifies the preference for clients). The order is now AES256,
2460 AES128, 3DES, ARCFOUR128.
2462 51. Small patch to Sieve code - explicitly set From: when generating an
2465 52. Exim crashed if a remote delivery caused a very long error message to be
2466 recorded - for instance if somebody sent an entire SpamAssassin report back
2467 as a large number of 550 error lines. This bug was coincidentally fixed by
2468 increasing the size of one of Exim's internal buffers (big_buffer) that
2469 happened as part of the Exiscan merge. However, to be on the safe side, I
2470 have made the code more robust (and fixed the comments that describe what
2473 53. Now that there can be additional text after "Completed" in log lines (if
2474 the queue_time_overall log selector is set), a one-byte patch to exigrep
2475 was needed to allow it to recognize "Completed" as not the last thing in
2478 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2479 patch that reportedly fixes this has been added. I am not expert enough to
2480 create a test for it. This is what the patch creator wrote:
2482 "I found a little strange behaviour of ldap code when working with
2483 Windows 2003 AD Domain, where users was placed in more than one
2484 Organization Units. When I tried to give exim partial DN, the exit code
2485 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2486 But simultaneously result of request was absolutely normal ldap result,
2487 so I produce this patch..."
2489 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2490 REFERENCE, so I have modified the code to exclude the patch when that macro
2493 55. Some experimental protocols are using DNS PTR records for new purposes. The
2494 keys for these records are domain names, not reversed IP addresses. The
2495 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2496 leaves it alone. Component reversal etc. now happens only for IP addresses.
2499 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2501 57. Double the size of the debug message buffer (to 2048) so that more of very
2502 long debug lines gets shown.
2504 58. The exicyclog utility now does better if the number of log files to keep
2505 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2507 59. Two changes related to the smtp_active_hostname option:
2509 (1) $smtp_active_hostname is now available as a variable.
2510 (2) The default for smtp_banner uses $smtp_active_hostname instead
2511 of $primary_hostname.
2513 60. The host_aton() function is supposed to be passed a string that is known
2514 to be a valid IP address. However, in the case of IPv6 addresses, it was
2515 not checking this. This is a hostage to fortune. Exim now panics and dies
2516 if the condition is not met. A case was found where this could be provoked
2517 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2518 components; fortuitously, this particular loophole had already been fixed
2519 by change 4.50/55 above.
2521 If there are any other similar loopholes, the new check in host_aton()
2522 itself should stop them being exploited. The report I received stated that
2523 data on the command line could provoke the exploit when Exim was running as
2524 exim, but did not say which command line option was involved. All I could
2525 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2526 running as the user.
2529 61. There was a buffer overflow vulnerability in the SPA authentication code
2530 (which came originally from the Samba project). I have added a test to the
2531 spa_base64_to_bits() function which I hope fixes it.
2534 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2535 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2537 63. The daemon start-up calls getloadavg() while still root for those OS that
2538 need the first call to be done as root, but it missed one case: when
2539 deliver_queue_load_max is set with deliver_drop_privilege. This is
2540 necessary for the benefit of the queue runner, because there is no re-exec
2541 when deliver_drop_privilege is set.
2543 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2544 This has been fixed.
2546 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2547 was in use, was not putting the data itself into the right store pool;
2548 consequently, it could be overwritten for a subsequent message in the same
2549 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2552 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2554 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2555 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2556 the function were treating the return as a boolean value, which happened to
2557 work because 0=false and not-0=true, but is not correct code.
2559 68. The host_aton() function was not handling scoped IPv6 addresses (those
2560 with, for example, "%eth0" on the end) correctly.
2562 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2563 negated items (that is, ~something) in unsigned ints. Some compilers
2564 apparently mutter when there is no cast.
2566 70. If an address verification called from an ACL failed, and did not produce a
2567 user-specific message (i.e. there was only a "system" message), nothing was
2568 put in $acl_verify_message. In this situation, it now puts the system
2571 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2572 session; change 4.31/43 added the unwanted input to the log line - except
2573 that it did not do this in the start of session case. It now does.
2575 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2576 This is wrong and can cause the other end to generate a synchronization
2577 error if it is another Exim or anything else that does the synchronization
2578 check. A QUIT command is no longer sent after a timeout.
2580 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2581 during host lookups.
2583 74. The defer_ok option of callout verification was not working if it was used
2584 when verifying addresses in header lines, that is, for this case:
2586 verify = header_sender/callout=defer_ok
2588 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2589 those file descriptors could be used for SMTP connections. If anything
2590 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2591 could be sent to the SMTP client, causing chaos. The daemon now opens
2592 stdin, stdout, and stderr to /dev/null when it puts itself into the
2595 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2596 log by default. The user can override this with suitable Perl magic.
2598 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2599 the log message when discard triggers, was not working for the DATA ACL or
2600 for the non-SMTP ACL.
2602 78. Error message wording change in sieve.c.
2604 79. If smtp_accept_max_per_host was set, the number of connections could be
2605 restricted to fewer than expected, because the daemon was trying to set up
2606 a new connection before checking whether the processes handling previous
2607 connections had finished. The check for completed processes is now done
2608 earlier. On busy systems, this bug wouldn't be noticed because something
2609 else would have woken the daemon, and it would have reaped the completed
2612 80. If a message was submitted locally by a user whose login name contained one
2613 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2614 It caused a spool format error. I have fixed the spool reading code. A
2615 related problem was that the "from" clause in the Received: line became
2616 illegal because of the space(s). It is now covered by ${quote_local_part.
2618 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2621 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2623 83. Patch from Sieve maintainer to fix unterminated string problem in
2624 "vacation" handling.
2626 84. Some minor changes to the Linux configuration files to help with other
2627 OS variants using glibc.
2629 85. One more patch for Sieve to update vacation handling to latest spec.
2632 ----------------------------------------------------
2633 See the note above about the 4.44 and 4.50 releases.
2634 ----------------------------------------------------
2640 1. Change 4.43/35 introduced a bug that caused file counts to be
2641 incorrectly computed when quota_filecount was set in an appendfile
2644 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2645 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2648 3. Give more explanation in the error message when the command for a transport
2649 filter fails to execute.
2651 4. There are several places where Exim runs a non-Exim command in a
2652 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2653 was being done only for the command run by the queryprogram router. It is
2654 now done for all such subprocesses. The other cases are: ${run, transport
2655 filters, and the commands run by the lmtp and pipe transports.
2657 5. Some older OS have a limit of 256 on the maximum number of file
2658 descriptors. Exim was using setrlimit() to set 1000 as a large value
2659 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2660 systems. I've change it so that if it can't get 1000, it tries for 256.
2662 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2663 was an oversight, and furthermore, ever since the addition of extra
2664 controls (e.g. 4.43/32), the checks on when to allow different forms of
2665 "control" were broken. There should now be diagnostics for all cases when a
2666 control that does not make sense is encountered.
2668 7. $recipients is now available in the predata ACL (oversight).
2670 8. Tidy the search cache before the fork to do a delivery from a message
2671 received from the command line. Otherwise the child will trigger a lookup
2672 failure and thereby defer the delivery if it tries to use (for example) a
2673 cached ldap connection that the parent has called unbind on.
2675 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2676 of $address_data from the recipient verification was clobbered by the
2677 sender verification.
2679 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2680 was its contents. (It was OK if the option was not defined at all.)
2682 11. A "Completed" log line is now written for messages that are removed from
2683 the spool by the -Mrm option.
2685 12. $host_address is now set to the target address during the checking of
2686 ignore_target_hosts.
2688 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2689 being passed; this would have caused $sender_host_name to have been used if
2690 matching the list had actually called for a host name (not very likely,
2691 since this list is usually IP addresses). A host name is now passed as
2694 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2695 code that specifies a non-restarting handler (typically sigaction() in
2696 modern systems) in an attempt to fix a rare and obscure crash bug.
2698 15. Narrowed the window for a race in the daemon that could cause it to ignore
2699 SIGCHLD signals. This is not a major problem, because they are used only to
2700 wake it up if nothing else does.
2702 16. A malformed maildirsize file could cause Exim to calculate negative values
2703 for the mailbox size or file count. Odd effects could occur as a result.
2704 The maildirsize information is now recalculated if the size or filecount
2707 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2708 support for a long time. Removed HAVE_SYS_VFS_H.
2710 18. Updated exipick to current release from John Jetmore.
2712 19. Allow an empty sender to be matched against a lookup in an address list.
2713 Previously the only cases considered were a regular expression, or an
2716 20. Exim went into a mad DNS lookup loop when doing a callout where the
2717 host was specified on the transport, if the DNS lookup yielded more than
2720 21. The RFC2047 encoding function was originally intended for short strings
2721 such as real names; it was not keeping to the 75-character limit for
2722 encoded words that the RFC imposes. It now respects the limit, and
2723 generates multiple encoded words if necessary. To be on the safe side, I
2724 have increased the buffer size for the ${rfc2047: expansion operator from
2727 22. Failure to deliver a bounce message always caused it to be frozen, even if
2728 there was an errors_to setting on the router. The errors_to setting is now
2731 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2732 canonical form (fully expanded) before being placed in
2733 $sender_host_address.
2735 24. Updated eximstats to version 1.33
2737 25. Include certificate and key file names in error message when GnuTLS fails
2738 to set them up, because the GnuTLS error message doesn't include the name
2739 of the failing file when there is a problem reading it.
2741 26. Expand error message when OpenSSL has problems setting up cert/key files.
2744 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2745 (this can affect the format of dates).
2747 28. exim_tidydb, when checking for the continued existence of a message for
2748 which it has found a message-specific retry record, was not finding
2749 messages that were in split spool directories. Consequently, it was
2750 deleting retry records that should have stayed in existence.
2752 29. eximstats updated to version 1.35
2753 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2754 1.35 - bugfix such that pie charts by volume are generated correctly
2756 30. The SPA authentication driver was not abandoning authentication and moving
2757 on to the next authenticator when an expansion was forced to fail,
2758 contradicting the general specification for all authenticators. Instead it
2759 was generating a temporary error. It now behaves as specified.
2761 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2762 (the order specifies the preference for clients). The order is now AES256,
2763 AES128, 3DES, ARCFOUR128.
2765 31. Small patch to Sieve code - explicitly set From: when generating an
2768 32. Exim crashed if a remote delivery caused a very long error message to be
2769 recorded - for instance if somebody sent an entire SpamAssassin report back
2770 as a large number of 550 error lines. This bug was coincidentally fixed by
2771 increasing the size of one of Exim's internal buffers (big_buffer) that
2772 happened as part of the Exiscan merge. However, to be on the safe side, I
2773 have made the code more robust (and fixed the comments that describe what
2776 33. Some experimental protocols are using DNS PTR records for new purposes. The
2777 keys for these records are domain names, not reversed IP addresses. The
2778 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2779 leaves it alone. Component reversal etc. now happens only for IP addresses.
2782 34. The host_aton() function is supposed to be passed a string that is known
2783 to be a valid IP address. However, in the case of IPv6 addresses, it was
2784 not checking this. This is a hostage to fortune. Exim now panics and dies
2785 if the condition is not met. A case was found where this could be provoked
2786 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2787 components; fortuitously, this particular loophole had already been fixed
2788 by change 4.50/55 or 4.44/33 above.
2790 If there are any other similar loopholes, the new check in host_aton()
2791 itself should stop them being exploited. The report I received stated that
2792 data on the command line could provoke the exploit when Exim was running as
2793 exim, but did not say which command line option was involved. All I could
2794 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2795 running as the user.
2798 35. There was a buffer overflow vulnerability in the SPA authentication code
2799 (which came originally from the Samba project). I have added a test to the
2800 spa_base64_to_bits() function which I hope fixes it.
2803 36. The daemon start-up calls getloadavg() while still root for those OS that
2804 need the first call to be done as root, but it missed one case: when
2805 deliver_queue_load_max is set with deliver_drop_privilege. This is
2806 necessary for the benefit of the queue runner, because there is no re-exec
2807 when deliver_drop_privilege is set.
2809 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2810 was in use, was not putting the data itself into the right store pool;
2811 consequently, it could be overwritten for a subsequent message in the same
2812 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2815 38. Sometimes the final signoff response after QUIT could fail to get
2816 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2817 < 0 before doing a fflush(). This bug looks as though it goes back to the
2818 introduction of TLS in release 3.20, but "sometimes" must have been rare
2819 because the tests only now provoked it.
2825 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2826 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2827 Now it returns an integer. A number of calls were still expecting a T/F
2828 return. Fortuitously, in all cases, the tests worked in OK situations,
2829 which is the norm. However, things would have gone wrong on any write
2830 failures on the smtp file descriptor. This function is used when sending
2831 messages over SMTP and also when doing verify callouts.
2833 2. When Exim is called to do synchronous delivery of a locally submitted
2834 message (the -odf or -odi options), it no longer closes stderr before doing
2837 3. Implemented the mua_wrapper option.
2839 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2841 5. Implemented the functions header_remove(), header_testname(),
2842 header_add_at_position(), and receive_remove_recipient(), and exported them
2845 6. If an ACL "warn" statement specified the addition of headers, Exim already
2846 inserted X-ACL-Warn: at the start if there was no header name. However, it
2847 was not making this test for the second and subsequent header lines if
2848 there were newlines in the string. This meant that an invalid header could
2849 be inserted if Exim was badly configured.
2851 7. Allow an ACL "warn" statement to add header lines at the start or after all
2852 the Received: headers, as well as at the end.
2854 8. Added the rcpt_4xx retry error code.
2856 9. Added postmaster_mailfrom=xxx to callout verification option.
2858 10. Added mailfrom=xxxx to the callout verification option, for verify=
2861 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2862 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2864 12. Inserted some casts to stop certain compilers warning when using pointer
2865 differences as field lengths or precisions in printf-type calls (mostly
2866 affecting debugging statements).
2868 13. Added optional readline() support for -be (dynamically loaded).
2870 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2871 same clock tick as a message's arrival, so that its received time was the
2872 same as the "first fail" time on the retry record, and that message
2873 remained on the queue past the ultimate address timeout, every queue runner
2874 would try a delivery (because it was past the ultimate address timeout) but
2875 after another failure, the ultimate address timeout, which should have then
2876 bounced the address, did not kick in. This was a "< instead of <=" error;
2877 in most cases the first failure would have been in the next clock tick
2878 after the received time, and all would be well.
2880 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2881 being recognized when the domain list was tested by the match_domain
2882 condition in an expansion string.
2884 16. Added the ${str2b64: operator.
2886 17. Exim was always calling setrlimit() to set a large limit for the number of
2887 processes, without checking whether the existing limit was already
2888 adequate. (It did check for the limit on file descriptors.) Furthermore,
2889 errors from getrlimit() and setrlimit() were being ignored. Now they are
2890 logged to the main and panic logs, but Exim does carry on, to try to do its
2891 job under whatever limits there are.
2893 18. Imported PCRE 5.0.
2895 19. Trivial typo in log message " temporarily refused connection" (the leading
2898 20. If the log selector return_path_on_delivery was set and an address was
2899 redirected to /dev/null, the delivery process crashed because it assumed
2900 that a return path would always be set for a "successful" delivery. In this
2901 case, the whole delivery is bypassed as an optimization, and therefore no
2904 21. Internal re-arrangement: the function for sending a challenge and reading
2905 a response while authentication was assuming a zero-terminated challenge
2906 string. It's now changed to take a pointer and a length, to allow for
2907 binary data in such strings.
2909 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2911 23. Exim was not respecting finduser_retries when seeking the login of the
2912 uid under which it was called; it was always trying 10 times. (The default
2913 setting of finduser_retries is zero.) Also, it was sleeping after the final
2914 failure, which is pointless.
2916 24. Implemented tls_on_connect_ports.
2918 25. Implemented acl_smtp_predata.
2920 26. If the domain in control=submission is set empty, Exim assumes that the
2921 authenticated id is a complete email address when it generates From: or
2922 Sender: header lines.
2924 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2925 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2926 chown and chgrp in /bin and hostname in /usr/bin.
2928 28. Exim was keeping the "process log" file open after each use, just as it
2929 does for the main log. This opens the possibility of it remaining open for
2930 long periods when the USR1 signal hits a daemon. Occasional processlog
2931 errors were reported, that could have been caused by this. Anyway, it seems
2932 much more sensible not to leave this file open at all, so that is what now
2935 29. The long-running daemon process does not normally write to the log once it
2936 has entered its main loop, and it closes the log before doing so. This is
2937 so that log files can straightforwardly be renamed and moved. However,
2938 there are a couple of unusual error situations where the daemon does write
2939 log entries, and I had neglected to close the log afterwards.
2941 30. The text of an SMTP error response that was received during a remote
2942 delivery was being truncated at 512 bytes. This is too short for some of
2943 the long messages that one sometimes sees. I've increased the limit to
2946 31. It is now possible to make retry rules that apply only when a message has a
2947 specific sender, in particular, an empty sender.
2949 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2950 it possible to be selective about when SMTP synchronization is enforced.
2952 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2954 32. Implemented hosts_connection_nolog.
2956 33. Added an ACL for QUIT.
2958 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2961 35. Added mailbox_size and mailbox_filecount to appendfile.
2963 36. Added control = no_multiline_responses to ACLs.
2965 37. There was a bug in the logic of the code that waits for the clock to tick
2966 in the case where the clock went backwards by a substantial amount such
2967 that the microsecond fraction of "now" was more than the microsecond
2968 fraction of "then" (but the whole seconds number was less).
2970 38. Added support for the libradius Radius client library this is found on
2971 FreeBSD (previously only the radiusclient library was supported).
2977 1. When certain lookups returned multiple values in the form name=value, the
2978 quoting of the values was not always being done properly. Specifically:
2979 (a) If the value started with a double quote, but contained no whitespace,
2981 (b) If the value contained whitespace other than a space character (i.e.
2982 tabs or newlines or carriage returns) it was not quoted.
2983 This fix has been applied to the mysql and pgsql lookups by writing a
2984 separate quoting function and calling it from the lookup code. The fix
2985 should probably also be applied to nisplus, ibase and oracle lookups, but
2986 since I cannot test any of those, I have not disturbed their existing code.
2988 2. A hit in the callout cache for a specific address caused a log line with no
2989 reason for rejecting RCPT. Now it says "Previous (cached) callout
2990 verification failure".
2992 3. There was an off-by-one bug in the queryprogram router. An over-long
2993 return line was truncated at 256 instead of 255 characters, thereby
2994 overflowing its buffer with the terminating zero. As well as fixing this, I
2995 have increased the buffer size to 1024 (and made a note to document this).
2997 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2998 when Exim is waiting for an SMTP response from a remote server, Exim
2999 restarts its select() call on the socket, thereby resetting its timeout.
3000 This is not a problem when such interrupts are rare. Somebody set up a cron
3001 job to run exiwhat every 2 minutes, which is less than the normal select()
3002 timeout (5 or 10 minutes). This meant that the select() timeout never
3003 kicked in because it was always reset. I have fixed this by comparing the
3004 time when an interrupt arrives with the time at the start of the first call
3005 to select(). If more time than the timeout has elapsed, the interrupt is
3006 treated as a timeout.
3008 5. Some internal re-factoring in preparation for the addition of Sieve
3009 extensions (by MH). In particular, the "personal" test is moved to a
3010 separate function, and given an option for scanning Cc: and Bcc: (which is
3011 not set for Exim filters).
3013 6. When Exim created an email address using the login of the caller as the
3014 local part (e.g. when creating a From: or Sender: header line), it was not
3015 quoting the local part when it contained special characters such as @.
3017 7. Installed new OpenBSD configuration files.
3019 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3020 try to make them clearer.
3022 9. Callout options, other than the timeout value, were being ignored when
3023 verifying sender addresses in header lines. For example, when using
3025 verify = header_sender/callout=no_cache
3027 the cache was (incorrectly) being used.
3029 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3030 only the Cygwin environment.
3032 11. When return_path_on_delivery was set as a log selector, if different remote
3033 addresses in the same message used different return paths and parallel
3034 remote delivery occurred, the wrong values would sometimes be logged.
3035 (Whenever a remote delivery process finished, the return path value from
3036 the most recently started remote delivery process was logged.)
3038 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3039 lines when AUTH and/or TLS are in use. This is the "received protocol"
3040 field. Exim used to use "asmtp" for authenticated SMTP, without any
3041 indication (in the protocol name) for TLS use. Now it follows the RFC and
3042 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3043 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3044 names appear in log lines as well as in Received: header lines.
3046 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3047 extensions, and comparison tests, and to fix some bugs.
3049 14. Changes to the "personal" filter test:
3051 (1) The test was buggy in that it was just doing the equivalent of
3052 "contains" tests on header lines. For example, if a user's address was
3053 anne@some.where, the "personal" test would incorrectly be true for
3055 To: susanne@some.where
3057 This test is now done by extracting each address from the header in turn,
3058 and checking the entire address. Other tests that are part of "personal"
3059 are now done using regular expressions (for example, to check local parts
3060 of addresses in From: header lines).
3062 (2) The list of non-personal local parts in From: addresses has been
3063 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3064 taken from the Sieve specification recommendations.
3066 (3) If the message contains any header line starting with "List-" it is
3067 treated as non-personal.
3069 (4) The test for "circular" in the Subject: header line has been removed
3070 because it now seems ill-conceived.
3072 15. Minor typos in src/EDITME comments corrected.
3074 16. Installed latest exipick from John Jetmore.
3076 17. If headers_add on a router specified a text string that was too long for
3077 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3078 of string_sprintf() is now avoided.
3080 18. $message_body_size was not set (it was always zero) when running the DATA
3081 ACL and the local_scan() function.
3083 19. For the "mail" command in an Exim filter, no default was being set for
3084 the once_repeat time, causing a random time value to be used if "once" was
3085 specified. (If the value happened to be <= 0, no repeat happened.) The
3086 default is now 0s, meaning "never repeat". The "vacation" command was OK
3087 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3088 (I found it when inspecting the code).
3090 20. There is now an overall timeout for performing a callout verification. It
3091 defaults to 4 times the callout timeout, which applies to individual SMTP
3092 commands during the callout. The overall timeout applies when there is more
3093 than one host that can be tried. The timeout is checked before trying the
3094 next host. This prevents very long delays if there are a large number of
3095 hosts and all are timing out (e.g. when the network connections are timing
3096 out). The value of the overall timeout can be changed by specifying an
3097 additional sub-option for "callout", called "maxwait". For example:
3099 verify = sender/callout=5s,maxwait=20s
3101 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3102 to the end before writing, but this should make it even safer).
3104 22. Exim was forgetting that it had advertised PIPELINING for the second and
3105 subsequent messages on an SMTP connection. It was also not resetting its
3106 memory on STARTTLS and an internal HELO.
3108 23. When Exim logs an SMTP synchronization error within a session, it now
3109 records whether PIPELINING has been advertised or not.
3111 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3112 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3113 rather than long int.
3115 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3117 26. Added the never_mail option to autoreply.
3123 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3124 crash if the getsockname() call failed; this can happen if a connection is
3125 closed very soon after it is established. The problem was simply in the
3126 order in which certain operations were done, causing Exim to try to write
3127 to the SMTP stream before it had set up the file descriptor. The bug has
3128 been fixed by making things happen in the correct order.
3134 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3135 before the connection was closed, thus losing the rejection response.
3137 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3138 some early Solaris releases, but causes trouble in current releases where
3139 socklen_t is defined.
3141 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3144 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3147 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3148 that releasing the top of it at the end releases what was used for sub-
3149 expansions (unless the block got too big). However, discard this block if
3150 the first thing is a variable or header, so that we can use its block when
3151 it is dynamic (useful for very large $message_headers, for example).
3153 6. Lookups now cache *every* query, not just the most recent. A new, separate
3154 store pool is used for this. It can be recovered when all lookup caches are
3155 flushed. Lookups now release memory at the end of their result strings.
3156 This has involved some general refactoring of the lookup sources.
3158 7. Some code has been added to the store_xxx() functions to reduce the amount
3159 of flapping under certain conditions.
3161 8. log_incoming_interface used to affect only the <= reception log lines. Now
3162 it causes the local interface and port to be added to several more SMTP log
3163 lines, for example "SMTP connection from", and rejection lines.
3165 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3167 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3169 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3170 could be overwritten at the end of the current message (or the start of a
3171 new message if it was set in a HELO ACL). The value is now preserved for
3172 the duration of the SMTP connection.
3174 12. If a transport had a headers_rewrite setting, and a matching header line
3175 contained an unqualified address, that address was qualified, even if it
3176 did not match any rewriting rules. The underlying bug was that the values
3177 of the flags that permit the existence of unqualified sender and recipient
3178 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3179 non-local messages, and by -bnq for local messages) were not being
3180 preserved with the message after it was received.
3182 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3183 "next input=" as part of the text comprising the host identity instead of
3184 the correct text. The code was using the same buffer for two different
3185 strings. However, depending on which order the printing function evaluated
3186 its arguments, the bug did not always show up. Under Linux, for example, my
3187 test suite worked just fine.
3189 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3190 doesn't work with some older versions of Perl. It has been changed to "my",
3191 which in any case is probably the better facility to use.
3193 15. A really picky compiler found some instances of statements for creating
3194 error messages that either had too many or two few arguments for the format
3197 16. The size of the buffer for calls to the DNS resolver has been increased
3198 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3199 for addresses that have a lot of PTR records. This alleviates a problem; it
3200 does not fully solve it.
3202 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3203 buffer now truncates the list and logs the incident, which is the same
3204 action as happens when Exim is looking up a host name and its aliases.
3205 Previously in this situation something unpredictable would happen;
3206 sometimes it was "internal error: store_reset failed".
3208 18. If a server dropped the connection unexpectedly when an Exim client was
3209 using GnuTLS and trying to read a response, the client delivery process
3210 crashed while trying to generate an error log message.
3212 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3213 string, for example:
3215 warn message = H1: something\nH2: something
3217 the text was added as a single header line from Exim's point of view
3218 though it ended up OK in the delivered message. However, searching for the
3219 second and subsequent header lines using $h_h2: did not work. This has been
3220 fixed. Similarly, if a system filter added multiple headers in this way,
3221 the routers could not see them.
3223 20. Expanded the error message when iplsearch is called with an invalid key to
3224 suggest using net-iplsearch in a host list.
3226 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3227 ACLs are no longer actually imposed (and a message to that effect is
3230 22. If a "gecos" field in a passwd entry contained escaped characters, in
3231 particular, if it contained a \" sequence, Exim got it wrong when building
3232 a From: or a Sender: header from that name. A second bug also caused
3233 incorrect handling when an unquoted " was present following a character
3234 that needed quoting.
3236 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3237 was not being matched caselessly.
3239 24. Arranged for all hyphens in the exim.8 source to be escaped with
3242 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3243 properly. Recipient callout cache records were still being keyed to include
3244 the sender, even when use_sender was set false. This led to far more
3245 callouts that were necessary. The sender is no longer included in the key
3246 when use_sender is false.
3248 26. Added "control = submission" modifier to ACLs.
3250 27. Added the ${base62d: operator to decode base 62 numbers.
3252 28. dnsdb lookups can now access SRV records.
3254 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3255 the configuration file.
3257 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3258 (-v) mode. This makes the output for a verbose queue run more intelligible.
3260 31. Added a use_postmaster feature to recipient callouts.
3262 32. Added the $body_zerocount variable, containing the number of binary zero
3263 bytes in the message body.
3265 33. The time of last modification of the "new" subdirectory is now used as the
3266 "mailbox time last read" when there is a quota error for a maildir
3269 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3271 35. Added +ignore_unknown as a special item in host lists.
3273 36. Code for decoding IPv6 addresses in host lists is now included, even if
3274 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3275 address was recognized as an IP address, but was then not correctly decoded
3276 into binary, causing unexpected and incorrect effects when compared with
3283 1. Very minor rewording of debugging text in manualroute to say "list of
3284 hosts" instead of "hostlist".
3286 2. If verify=header_syntax was set, and a header line with an unqualified
3287 address (no domain) and a large number of spaces between the end of the
3288 name and the colon was received, the reception process suffered a buffer
3289 overflow, and (when I tested it) crashed. This was caused by some obsolete
3290 code that should have been removed. The fix is to remove it!
3292 3. When running in the test harness, delay a bit after writing a bounce
3293 message to get a bit more predictability in the log output.
3295 4. Added a call to search_tidyup() just before forking a reception process. In
3296 theory, someone could use a lookup in the expansion of smtp_accept_max_
3297 per_host which, without the tidyup, could leave open a database connection.
3299 5. Added the variables $recipient_data and $sender_data which get set from a
3300 lookup success in an ACL "recipients" or "senders" condition, or a router
3301 "senders" option, similar to $domain_data and $local_part_data.
3303 6. Moved the writing of debug_print from before to after the "senders" test
3306 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3307 problems for message scanning, either using a data ACL, or using
3308 local_scan() because the Received: header was not generated till after they
3309 were called (in order to set the time as the time of reception completion).
3310 I have revised the way this works. The header is now generated after the
3311 body is received, but before the ACL or local_scan() are called. After they
3312 are run, the timestamp in the header is updated.
3318 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3319 before starting a queue runner without re-exec. This happened only when
3320 deliver_drop_privilege was set or when the Exim user was set to root. The
3321 effect of the bug was that timeouts during subsequent deliveries caused
3322 crashes instead of being properly handled. The handler is now left at its
3323 default (and expected) setting.
3325 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3326 message, again when deliver_drop_privilege is set or Exim is run as root.
3327 The bug described in (1) was not present in this case, but the tidying up
3328 of the other signals was missing. I have made the two cases consistent.
3330 3. The ignore_target_hosts setting on a manualroute router was being ignored
3331 for hosts that were looked up using the /MX notation.
3333 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3336 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3337 operated on the sender address. After changing the $sender_address to <>
3338 for the sender address verify, Exim was re-instated it as the original
3339 (before rewriting) address, but remembering that it had rewritten it, so it
3340 wasn't rewriting it again. This bug also had the effect of breaking the
3341 sender address verification caching when the sender address was rewritten.
3343 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3344 This has been changed so that if the ip literal address matches
3345 ignore_target_hosts, the router declines.
3347 7. Added expansion conditions match_domain, match_address, and match_local_
3348 part (NOT match_host).
3350 8. The placeholder for the Received: header didn't have a length field set.
3352 9. Added code to Exim itself and to exim_lock to test for a specific race
3353 condition that could lead to file corruption when using MBX delivery. The
3354 issue is with the lockfile that is created in /tmp. If this file is removed
3355 after a process has opened it but before that process has acquired a lock,
3356 there is the potential for a second process to recreate the file and also
3357 acquire a lock. This could lead to two Exim processes writing to the file
3358 at the same time. The added code performs the same test as UW imapd; it
3359 checks after acquiring the lock that its file descriptor still refers to
3360 the same named file.
3362 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3363 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3366 11. Added the smtp_active_hostname option. If used, this will typically be made
3367 to depend on the incoming interface address. Because $interface_address is
3368 not set up until the daemon has forked a reception process, error responses
3369 that can happen earlier (such as "too many connections") no longer contain
3372 12. If an expansion in a condition on a "warn" statement fails because a lookup
3373 defers, the "warn" statement is abandoned, and the next ACL statement is
3374 processed. Previously this caused the whole ACL to be aborted.
3376 13. Added the iplsearch lookup type.
3378 14. Added ident_timeout as a log selector.
3380 15. Added tls_certificate_verified as a log selector.
3382 16. Added a global option tls_require_ciphers (compare the smtp transport
3383 option of the same name). This controls incoming TLS connections.
3385 17. I finally figured out how to make tls_require_ciphers do a similar thing
3386 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3387 before starting the TLS session.
3389 18. Tabs are now shown as \t in -bP output.
3391 19. If the log selector return_path_on_delivery was set, Exim crashed when
3392 bouncing a message because it had too many Received: header lines.
3394 20. If two routers both had headers_remove settings, and the first one included
3395 a superfluous trailing colon, the final name in the first list and the
3396 first name in the second list were incorrectly joined into one item (with a
3397 colon in the middle).
3403 1. Added -C and -D options to the exinext utility, mainly to make it easier
3404 to include in the automated testing, but these could be helpful when
3405 multiple configurations are in use.
3407 2. The exinext utility was not formatting the output nicely when there was
3408 an alternate port involved in the retry record key, nor when there was a
3409 message id as well (for retries that were specific to a specific message
3410 and a specific host). It was also confused by IPv6 addresses, because of
3411 the additional colons they contain. I have fixed the IPv4 problem, and
3412 patched it up to do a reasonable job for IPv6.
3414 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3415 delivery, the log line now contains "pipelined" if PIPELINING was used.
3417 4. An SMTP transport process used to panic and die if the bind() call to set
3418 an explicit outgoing interface failed. This has been changed; it is now
3419 treated in the same way as a connect() failure.
3421 5. A reference to $sender_host_name in the part of a conditional expansion
3422 that was being skipped was still causing a DNS lookup. This no longer
3425 6. The def: expansion condition was not recognizing references to header lines
3426 that used bh_ and bheader_.
3428 7. Added the _cache feature to named lists.
3430 8. The code for checking quota_filecount in the appendfile transport was
3431 allowing one more file than it should have been.
3433 9. For compatibility with Sendmail, the command line option
3441 and sets the incoming protocol and host name (for trusted callers). The
3442 host name and its colon can be omitted when only the protocol is to be set.
3443 Note the Exim already has two private options, -pd and -ps, that refer to
3444 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3445 "s", but I don't think that's a major issue.
3447 10. A number of refactoring changes to the code, none of which should affect
3450 (a) The number of logging options was getting close to filling up the
3451 32-bit word that was used as a bit map. I have split them into two classes:
3452 those that are passed in the argument to log_write(), and those that are
3453 only ever tested independently outside of that function. These are now in
3454 separate 32-bit words, so there is plenty of room for expansion again.
3455 There is no change in the user interface or the logging behaviour.
3457 (b) When building, for example, log lines, the code previously used a
3458 macro that called string_cat() twice, in order to add two strings. This is
3459 not really sufficiently general. Furthermore, there was one instance where
3460 it was actually wrong because one of the argument was used twice, and in
3461 one call a function was used. (As it happened, calling the function twice
3462 did not affect the overall behaviour.) The macro has been replaced by a
3463 function that can join an arbitrary number of extra strings onto a growing
3466 (c) The code for expansion conditions now uses a table and a binary chop
3467 instead of a serial search (which was left over from when there were very
3468 few conditions). Also, it now recognizes conditions like "pam" even when
3469 the relevant support is not compiled in: a suitably worded error message is
3470 given if an attempt is made to use such a condition.
3472 11. Added ${time_interval:xxxxx}.
3474 12. A bug was causing one of the ddress fields not to be passed back correctly
3475 from remote delivery subprocesses. The field in question was not being
3476 subsequently used, so this caused to problems in practice.
3478 13. Added new log selectors queue_time and deliver_time.
3480 14. Might have fixed a bug in maildirsizefile handling that threw up
3481 "unexpected character" debug warnings, and recalculated the data
3482 unnecessarily. In any case, I expanded the warning message to give more
3485 15. Added the message "Restricted characters in address" to the statements in
3486 the default ACL that block characters like @ and % in local parts.
3488 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3489 Three changes have been made:
3491 (a) There was a serious bug; a negative response to MAIL caused the whole
3492 recipient domain to be cached as invalid, thereby blocking all messages
3493 to all local parts at the same domain, from all senders. This bug has
3494 been fixed. The domain is no longer cached after a negative response to
3495 MAIL if the sender used is not empty.
3497 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3500 (c) A new callout option, "use_sender" has been added for people who want
3501 the modified behaviour.
3507 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3510 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3511 indeed breaks things for older releases.
3513 3. Added additional logging to the case where there is a problem reading data
3514 from a filter that is running in a subprocess using a pipe, in order to
3515 try to track down a specific problem.
3517 4. Testing facility fudge: when running in the test harness and attempting
3518 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3519 getting "No route to host". Convert this to a timeout.
3521 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3524 6. Some OS don't have socklen_t but use size_t instead. This affects the
3525 fifth argument of getsockopt() amongst other things. This is now
3526 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3527 can be set for individual OS. I have set it for SunOS5, OSF1, and
3528 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3529 some earlier ones do not.
3531 7. Change 4.30/15 was not doing the test caselessly.
3533 8. The standard form for an IPv6 address literal was being rejected by address
3534 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3535 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3536 this, as well as the form without the "IPv6" on the front (but only when
3537 address literals are enabled, of course).
3539 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3541 10. Exim crashed if a message with an empty sender address specified by -f
3542 encountered a router with an errors_to setting. This could be provoked only
3543 by a command such as
3547 where an empty string was supplied; "<>" did not hit this bug.
3549 11. Installed PCRE release 4.5.
3551 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3552 remained set. It is now erased.
3554 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3555 times from message ids (which are base 36 rather than the normal 62).
3557 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3558 were being counted as actual protocol errors, and logged if the log
3559 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3560 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3561 and DATA following a set of rejected RCPTs do not count as protocol errors.
3562 In other words, Exim assumes they were pipelined, though this may not
3563 actually be the case. Of course, in all cases the client gets an
3564 appropriate error code.
3566 15. If a lookup fails in an ACL condition, a message about the failure may
3567 be available; it is used if testing the ACL cannot continue, because most
3568 such messages specify what the cause of the deferral is. However, some
3569 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3570 that caused an old message to be retained and used if a later statement
3571 caused a defer, replacing the real cause of the deferral.
3573 16. If an IP address had so many PTR records that the DNS lookup buffer
3574 was not large enough to hold them, Exim could crash while trying to process
3575 the truncated data. It now detects and logs this case.
3577 17. Further to 4.21/58, another change has been made: if (and only if) the
3578 first line of a message (the first header line) ends with CRLF, a bare LF
3579 in a subsequent header line has a space inserted after it, so as not to
3580 terminate the header.
3582 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3583 unnecessarily, used atoi() instead of strtol(), and didn't check the
3584 termination when getting file sizes from file names by regex.
3586 19. Completely re-implemented the support for maildirsize files, in the light
3587 of a number of problems with the previous contributed implementation
3588 (4.30/29). In particular:
3590 . If the quota is zero, the maildirsize file is maintained, but no quota is
3593 . If the maildir directory does not exist, it is created before any attempt
3594 to write a maildirsize file.
3596 . The quota value in the file is just a cache; if the quota is changed in
3597 the transport, the new value overrides.
3599 . A regular expression is available for excluding directories from the
3602 20. The autoreply transport checks the characters in options that define the
3603 message's headers; it allows continued headers, but it was checking with
3604 isspace() after an embedded newline instead of explicitly looking for a
3607 21. If all the "regular" hosts to which an address was routed had passed their
3608 expiry times, and had not reached their retry times, the address was
3609 bounced, even if fallback hosts were defined. Now Exim should go on to try
3612 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3613 equivalent code in the SMTP transport. Some hosts send humungous responses
3614 to HELO/EHLO, more than 1024 it seems.
3616 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3617 gives compiler warnings in some environments. I've now done it "properly",
3620 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3621 (because of problems with the built-in one) was declared to return uschar *
3622 instead of char *, causing compiler failure.
3624 25. Fixed a file descriptor leak when processing alias/forward files.
3626 26. Fixed a minor format string issue in dbfn.c.
3628 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3630 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3631 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3634 29. When a sender address is verified, it is cached, to save repeating the test
3635 when there is more than one recipient in a message. However, when the
3636 verification involves a callout, it is possible for different callout
3637 options to be set for different recipients. It is too complicated to keep
3638 track of this in the cache, so now Exim always runs a verification when a
3639 callout is required, relying on the callout cache for the optimization.
3640 The overhead is duplication of the address routing, but this should not be
3643 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3644 to be verified with callout=postmaster, and the main callout worked but the
3645 postmaster check failed, the verification correctly failed. However, if a
3646 subsequent RCPT command asked for sender verification *without* the
3647 postmaster check, incorrect caching caused this verification also to fail,
3650 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3651 it was not caching the DNS options (qualify_single, search_parents) that
3652 were used when the lookup failed. A subsequent lookup with different
3653 options therefore always gave the same answer, though there were cases
3654 where it should not have. (Example: a "domains = !$mx_any" option on a
3655 dnslookup router: the "domains" option is always processed without any
3656 widening, but the router might have qualify_single set.) Now Exim uses the
3657 cached value only when the same options are set.
3659 32. Added John Jetmore's "exipick" utility to the distribution.
3661 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3662 than a timeout (e.g. a certificate is required, and is not provided), an
3663 Exim server now closes the connection immediately. Previously it waited for
3664 the client to close - but if the client is SSL, it seems that they each
3665 wait for each other, leading to a delay before one of them times out.
3667 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3668 maintained 0.8.x compatibility because I don't think many are using it, and
3669 it is clearly obsolete.
3671 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3674 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3675 client certificate was expired. A simple patch fixes this, though I don't
3676 understand the full logic of why the verify callback is called multiple
3679 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3680 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3681 which causes problems with some clients (such as the Certicom SSL Plus
3682 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3683 disables the coutermeasure allowing Eudora to connect."
3685 38. Exim was not checking that a write() to a log file succeeded. This could
3686 lead to Bad Things if a log got too big, in particular if it hit a file
3687 size limit. Exim now panics and dies if it cannot write to a log file, just
3688 as it does if it cannot open a log file.
3690 39. Modified OS/Makefile-Linux so that it now contains
3692 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3694 The two -D definitions ensure that Exim is compiled with large file
3695 support, which makes it possible to handle log files that are bigger than
3698 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3699 instance) a domain was checked against a named list that involved a lookup,
3700 causing $domain_data to be set, then another domain was checked against the
3701 same list, then the first domain was re-checked, the value of $domain_data
3702 after the final check could be wrong. In particular, if the second check
3703 failed, it could be set empty. This bug probably also applied to
3706 41. The strip_trailing_dot option was not being applied to the address given
3707 with the -f command-line option.
3709 42. The code for reading a message's header from the spool was incrementing
3710 $received_count, but never initializing it. This meant that the value was
3711 incorrect (doubled) while delivering a message in the same process in which
3712 it was received. In the most common configuration of Exim, this never
3713 happens - a fresh exec is done - but it can happen when
3714 deliver_drop_privilege is set.
3716 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3717 it now includes up to 150 characters of the unexpected data in the log
3720 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3721 and building data strings. The size of both of these buffers was 10 000
3722 bytes - far larger than anybody would *ever* want, thought I. Needless to
3723 say, somebody hit the limit. I have increased the maximum line length to
3724 20 000 and the maximum data length of concatenated lines to 100 000. I have
3725 also fixed two bugs, because there was no checking on these buffers. Tsk,
3726 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3727 buffer is too small.
3729 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3730 lsearch lookups. Now it does.
3732 46. When parsing a route_list item in a manualroute router, a fixed-length
3733 buffer was used for the list of hosts. I made this 1024 bytes long,
3734 thinking that nobody would ever have a list of hosts that long. Wrong.
3735 Somebody had a whole pile of complicated expansion conditions, and the
3736 string was silently truncated, leading to an expansion error. It turns out
3737 that it is easier to change to an unlimited length (owing to other changes
3738 that have happened since this code was originally written) than to build
3739 structure for giving a limitation error. The length of the item that
3740 expands into the list of hosts is now unlimited.
3742 47. The lsearch lookup could not handle data where the length of text line was
3743 more than 4095 characters. Such lines were truncated, leading to shortened
3744 data being returned. It should now handle lines of any length.
3746 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3747 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3750 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3751 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3752 no longer happens in comment lines. A list of replacements is now placed
3753 at the head of all of the source files, except those whose only change is
3754 to replace PERL_COMMAND in the very first #! line.
3756 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3757 messages on the queue, with a bottom-up merge sort, using code contributed
3758 by Michael Haardt. This should make operations like -bp somewhat faster on
3759 large queues. It won't affect queue runners, except when queue_run_in_order
3762 51. Installed eximstats 1.31 in the distribution.
3764 52. Added support for SRV lookups to the dnslookup router.
3766 53. If an ACL referred to $message_body or $message_body_end, the value was not
3767 reset for any messages that followed in the same SMTP session.
3769 54. The store-handling optimization for building very long strings was not
3770 differentiating between the different store pools. I don't think this
3771 actually made any difference in practice, but I've tidied it.
3773 55. While running the routers to verify a sender address, $sender_address
3774 was still set to the sender address. This is wrong, because when routing to
3775 send a bounce to the sender, it would be empty. Therefore, I have changed
3776 it so that, while verifying a sender address, $sender_address is set to <>.
3777 (There is no change to what happens when verifying a recipient address.)
3779 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3780 target A or AAAA records (if not already returned) without resetting the
3781 qualify_single or search_parents options of the DNS resolver. These are
3782 inappropriate in this case because the targets of MX and SRV records must
3783 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3784 target that, when qualified, matched something in the local domain. These
3785 two options are now turned off when doing these lookups.
3787 57. It seems that at least some releases of Reiserfs (which does not have the
3788 concept of a fixed number of inodes) returns zero and not -1 for the
3789 number of available inodes. This interacted badly with check_spool_inodes,
3790 which assumed that -1 was the "no such thing" setting. What I have done is
3791 to check that the total number of inodes is greater than zero before doing
3792 the test of how many are available.
3794 58. When a "warn" ACL statement has a log_message modifier, the message is
3795 remembered, and not repeated. This is to avoid a lot of repetition when a
3796 message has many recipients that cause the same warning to be written.
3797 Howewer, Exim was preserving the list of already written lines for an
3798 entire SMTP session, which doesn't seem right. The memory is now reset if a
3799 new message is started.
3801 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3802 debugging output unless log_rewrite was also set.
3804 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3805 of (int)(handle) when we know that handle contains (void *)(-1).
3807 61. The Exim daemon panic-logs an error return when it closes the incoming
3808 connection. However "connection reset by peer" seems to be common, and
3809 isn't really an error worthy of noting specially, so that particular error
3812 62. When Exim is trying to find all the local interfaces, it used to panic and
3813 die if the ioctl to get the interface flags failed. However, it seems that
3814 on at least one OS (Solaris 9) it is possible to have an interface that is
3815 included in the list of interfaces, but for which you get a failure error
3816 for this call. This happens when the interface is not "plumbed" into a
3817 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3818 failure of the "get flags" call assumes that the interface is down.
3820 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3821 makes life easier for people who are doing arithmetic on fields extracted
3822 from dates, where you often get leading zeros that should not be
3823 interpreted as octal.
3825 64. Added qualify_domain to the redirect router, to override the global
3828 65. If a pathologically long header line contained very many addresses (the
3829 report of this problem mentioned 10 000) and each of them was rewritten,
3830 Exim could use up a very large amount of memory. (It kept on making new
3831 copies of the header line as it rewrote, and never released the old ones.)
3832 At the expense of a bit more processing, the header rewriting function has
3833 been changed so that it no longer eats memory in this way.
3835 66. The generation of the Received: header has been moved from the time that a
3836 message starts to be received, to the time that it finishes. The timestamp
3837 in the Received: header should now be very close to that of the <= log
3838 line. There are two side-effects of this change:
3840 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3841 logged header lines no longer include the local Received: line, because
3842 it has not yet been created. The same applies to a copy of the message
3843 that is returned to a non-SMTP sender when a message is rejected.
3845 (b) When a filter file is tested using -bf, no additional Received: header
3846 is added to the test message. After some thought, I decided that this
3849 This change does not affect the value of $received_for. It is still set
3850 after address rewriting, but before local_scan() is called.
3852 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3854 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3855 gave an unhelpful panic error message, and a defer error. I have managed to
3856 change this behaviour so that it now rejects any supplied certificate,
3857 which seems right, as the list of acceptable certificates is empty.
3859 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3860 gave an unhelpful defer error. I have not managed to make this reject any
3861 supplied certificates, but the error message it gives is "no certificate
3862 supplied", which is not helpful.
3864 70. exigrep's output now also includes lines that are not associated with any
3865 message, but which match the given pattern. Implemented by a patch from
3866 Martin Sluka, which also tidied up the Perl a bit.
3868 71. Recipient callout verification, like sender verification, was using <> in
3869 the MAIL FROM command. This isn't really the right thing, since the actual
3870 sender may affect whether the remote host accepts the recipient or not. I
3871 have changed it to use the actual sender in the callout; this means that
3872 the cache record is now keyed on a recipient/sender pair, not just the
3873 recipient address. There doesn't seem to be a real danger of callout loops,
3874 since a callout by the remote host to check the sender would use <>.
3875 [SEE ABOVE: changed after hitting problems.]
3877 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3878 temporary errors. However, in the case of such a code being given after
3879 the end of a data transmission (i.e. after ".") Exim was failing to write
3880 a retry record for the message. (Yes, there was some broken host that was
3881 actually sending 8xx at this point.)
3883 73. An unknown lookup type in a host list could cause Exim to panic-die when
3884 the list was checked. (An example that provoked this was putting <; in the
3885 middle of a list instead of at the start.) If this happened during a DATA
3886 ACL check, a -D file could be left lying around. This kind of configuration
3887 error no longer causes Exim to die; instead it causes a defer errror. The
3888 incident is still logged to the main and panic logs.
3890 74. Buglet left over from Exim 3 conversion. The message "too many messages
3891 in one connection" was written to the rejectlog but not the mainlog, except
3892 when address rewriting (yes!) was being logged.
3894 75. Added write_rejectlog option.
3896 76. When a system filter was run not as root (that is, when system_filter_user
3897 was set), the values of the $n variables were not being returned to the
3898 main process; thus, they were not subsequently available in the $sn
3901 77. Added +return_path_on_delivery log selector.
3903 78. A connection timeout was being treated differently from recipients deferred
3904 when testing hosts_max_try with a message that was older than the host's
3905 retry timeout. (The host should not be counted, thus allowing all hosts to
3906 be tried at least once before bouncing.) This may have been the cause of an
3907 occasionally reported bug whereby a message would remain on the queue
3908 longer than the retry timeout, but would be bounced if a delivery was
3909 forced. I say "may" because I never totally pinned down the problem;
3910 setting up timeout/retry tests is difficult. See also the next item.
3912 79. The ultimate address timeout was not being applied to errors that involved
3913 a combination of host plus message (for example, a timeout on a MAIL
3914 command). When an address resolved to a number of possible hosts, and they
3915 were not all tried for each delivery (e.g. because of hosts_max_try), a
3916 message could remain on the queue longer than the retry timeout.
3918 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3921 81. Fixed an obscure SMTP outgoing bug which required at least the following
3922 conditions: (a) there was another message waiting for the same server;
3923 (b) the server returned 5xx to all RCPT commands in the first message so
3924 that the message was not completed; (c) the server dropped the connection
3925 or gave a negative response to the RSET that Exim sends to abort the
3926 transaction. The observed case was a dropped connection after DATA that had
3927 been sent in pipelining mode. That is, the server had advertised PIPELINING
3928 but was not implementing it correctly. The effect of the bug was incorrect
3929 behaviour, such as trying another host, and this could lead to a crash.
3935 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3936 and daemon.c were passed as pointers to ints; they should have been
3937 pointers to socklen_t variables (which are typically unsigned ints).
3939 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3942 3. Fixed a really odd bug that affected only the testing scheme; patching a
3943 certain fixed string in the binary changed the value of another string that
3944 happened to be identical to the end of the original first string.
3946 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3947 name", it returns that address as the IP address. On some operating
3948 systems (e.g. Solaris), it also passes back the IP address string as the
3949 "host name". However, on others (e.g. Linux), it passes back an empty
3950 string. Exim wasn't checking for this, and was changing the host name to an
3951 empty string, assuming it had been canonicized.
3953 5. Although rare, it is permitted to have more than one PTR record for a given
3954 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3955 all the names associated with an address, because they do in Solaris.
3956 However, it seems that they do not in Linux for data that comes from the
3957 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3958 I found this out when I moved to a new Linux workstation and tried to run
3959 the Exim test suite.
3961 To get round this problem I have changed the code so that it now does its
3962 own call to the DNS to look up PTR records when searching for a host name.
3963 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3964 addresses that are only in /etc/hosts are still found.
3966 This behaviour is, however, controlled by an option called host_lookup_
3967 order, which defaults to "bydns:byaddr". If people want to use the other
3968 order, or indeed, just use one or the other means of lookup, they can
3969 specify it in this variable.
3971 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3972 some operating systems, this comes back from gethostbyaddr() as an empty
3973 string, and this is what Exim used to test for. However, it seems that in
3974 other systems, "." is yielded. Exim now tests for this case too.
3976 7. The values of check_spool_space and check_log_space are now held internally
3977 as a number of kilobytes instead of an absolute number of bytes. If a
3978 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3979 kilobyte. This means that much larger values can be stored.
3981 8. Exim monitor: an attempt to get the action menu when not actually pointing
3982 at a message produces an empty menu entitled "No message selected". This
3983 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3984 no entries in it ("Shell widget menu has zero width and/or height"). So I
3985 have added a single, blank menu entry in this case.
3987 9. Added ${quote_local_part.
3989 10. MIME decoding is now applied to the contents of Subject: header lines when
3992 11. Now that a reference to $sender_host_address automatically causes a reverse
3993 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3994 host lookup before query-style lookups in lists that might use this
3995 variable. This has therefore been abolished, and the "net-" prefix is no
3996 longer necessary for query-style lookups.
3998 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3999 have been a typo for LFLAGS, so it has been changed.
4001 13. The install script calls Exim with "-C /dev/null" in order to find the
4002 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4003 to be output. Howeve, since Exim outputs its version number before the
4004 error, it didn't break the script. It just looked ugly. I fixed this by
4005 always allowing "-C /dev/null" if the caller is root.
4007 14. Ignore overlarge ACL variable number when reading spool file - insurance
4008 against a later release with more variables having written the file.
4010 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4011 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4012 this, as well as the form without the "IPv6" on the front.
4014 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4015 OS/Makefile-Darwin file.
4017 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4018 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4020 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4021 variable. However, some OS use ints for inodes. I've added cast to long int
4022 to get rid of the compiler warning.
4024 19. I had forgotten to lock out "/../" in configuration file names when
4025 ALT_CONFIG_PREFIX was set.
4027 20. Routers used for verification do not need to specify transports. However,
4028 if such a router generated a host list, and callout was configured, Exim
4029 crashed, because it could not find a port number from the (non-existent)
4030 transport. It now assumes port 25 in this circumstance.
4032 21. Added the -t option to exigrep.
4034 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4035 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4036 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4037 them is set, LOOKUP_LSEARCH is forced.
4039 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4041 24. Added sender and host information to the "rejected by local_scan()" log
4042 line; previously there was no indication of these.
4044 25. Added .include_if_exists.
4046 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4047 Linux. It turns out that not all file systems support this. Apparently some
4048 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4049 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4050 not supported on the file descriptor, is now ignored when Exim is trying to
4051 sync a directory. This applies only to Linux.
4053 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4055 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4056 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4059 29. Added support for maildirsize files from supplied patch (modified a bit).
4061 30. The use of :fail: followed by an empty string could lead Exim to respond to
4062 sender verification failures with (e.g.):
4064 550 Verification failed for <xxx>
4065 550 Sender verify failed
4067 where the first response line was missing the '-' that indicates it is not
4068 the final line of the response.
4070 31. The loop for finding the name of the user that called Exim had a hardwired
4071 limit of 10; it now uses the value of finduser_retries, which is used for
4072 all other user lookups.
4074 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4077 33. Exim was neglecting to zero errno before one call of strtol() when
4078 expanding a string and expecting an integer value. On some systems this
4079 resulted in spurious "integer overflow" errors. Also, it was casting the
4080 result into an int without checking.
4082 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4083 did not work. The code looks as if it has *never* worked, though it appears
4084 to have been documented since at least releast 1.62. I have made it work.
4086 35. The "timeout_DNS" error in retry rules, also documented since at least
4087 1.62, also never worked. As it isn't clear exactly what this means, and
4088 clearly it isn't a major issue, I have abolished the feature by treating it
4089 as "timeout", and writing a warning to the main and panic logs.
4091 36. The display of retry rules for -brt wasn't always showing the error code
4094 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4095 timeout_connect_A, timeout_connect_MX.
4097 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4098 to the empty sender.
4100 39. The daemon was not analysing the content of -oX till after it had closed
4101 stderr and disconnected from the controlling terminal. This meant that any
4102 syntax errors were only noted on the panic log, and the return code from
4103 the command was 0. By re-arranging the code a little, I've made the
4104 decoding happen first, so such errors now appear on stderr, and the return
4105 code is 1. However, the actual setting up of the sockets still happens in
4106 the disconnected process, so errors there are still only recorded on the
4109 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4110 connections (as happens on some IP stacks) was logged at start up time as
4111 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4112 it from "IPv6 and IPv4", which means that two separate sockets are being
4115 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4116 says whether AF_INET or AF_INET6 was passed as an argument.
4118 42. Exiwhat output was messed up when time zones were included in log
4121 43. Exiwhat now gives more information about the daemon's listening ports,
4122 and whether -tls-on-connect was used.
4124 44. The "port" option of the smtp transport is now expanded.
4126 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4127 silently ignored. Now an error message is written to the main and panic
4130 46. There's a new ACL modifier called "logwrite" which writes to a log file
4131 as soon as it is encountered.
4133 47. Added $local_user_uid and $local_user_gid at routing time.
4135 48. Exim crashed when trying to verify a sender address that was being
4138 49. Exim was recognizing only a space character after ".include". It now also
4139 recognizes a tab character.
4141 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4142 extracting the relevant information from the specification. The man page no
4143 longer contains scrambled data for the -d option, and I've added a section
4144 at the front about calling Exim under different names.
4146 51. Added "extra_headers" argument to the "mail" command in filter files.
4148 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4151 53. Installed eximstats 1.29.
4153 54. Added transport_filter_timeout as a generic transport option.
4155 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4156 Cc: header lines. This was required by RFC 822, but it not required by RFC
4159 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4160 incoming messages that did not have them. Now it does so only if the
4161 message originates locally, that is, if there is no associated remote host
4162 address. When Resent- header lines are present, this applies to the Resent-
4163 lines rather than the non-Resent- lines.
4165 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4166 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4168 58. Messages for configuration errors now include the name of the main
4169 configuration file - useful now that there may be more than one file in a
4170 list (.included file names were always shown).
4172 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4173 for those rare installations that do not start the daemon as root or run it
4174 setuid root. I've cut out the call to initgroups() if the daemon is not
4177 60. The Exim user and group can now be bound into the binary as text strings
4178 that are looked up at the start of Exim's processing.
4180 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4182 62. Added $mailstore_basename variable.
4184 63. Installed patch to sieve.c from Michael Haardt.
4186 64. When Exim failed to open the panic log after failing to open the main log,
4187 the original message it was trying to log was written to stderr and debug
4188 output, but if they were not available (the usual case in production), it
4189 was lost. Now it is written to syslog before the two lines that record the
4190 failures to open the logs.
4192 65. Users' Exim filters run in subprocesses under the user's uid. It is
4193 possible for a "deliver" command or an alias in a "personal" command to
4194 provoke an address rewrite. If logging of address rewriting is configured,
4195 this fails because the process is not running as root or exim. There may be
4196 a better way of dealing with this, but for the moment (because 4.30 needs
4197 to be released), I have disabled address rewrite logging when running a
4198 filter in a non-root, non-exim process.
4204 1. The buildconfig auxiliary program wasn't quoting the value set for
4205 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4206 not defined. This bug was masked in 4.22 by the effect that was fixed in
4209 2. Some messages that were rejected after a message id was allocated were
4210 shown as "incomplete" by exigrep. It no longer does this for messages that
4211 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4213 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4214 have allow_domain_literals set, the ID did not get logged in the <= line.
4215 Domain literals are now always recognized in Message-ID: header lines.
4217 4. The first argument for a ${extract expansion item is the key name or field
4218 number. Leading and trailing spaces in this item were not being ignored,
4219 causing some misleading effects.
4221 5. When deliver_drop_privilege was set, single queue runner processes started
4222 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4223 same command in the process it spins off) were not dropping privilege.
4225 6. When the daemon running as "exim" started a queue runner, it always
4226 re-executed Exim in the spun-off process. This is a waste of effort when
4227 deliver_drop_privilege is set. The new process now just calls the
4228 queue-runner function directly.
4234 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4237 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4238 ignored. Though the use of -oP was forcing the writing of a pid file, it
4239 was always written to the default place.
4241 3. If the message "no IP address found for host xxxx" is generated during
4242 incoming verification, it is now followed by identification of the incoming
4243 connection (so you can more easily find what provoked it).
4245 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4247 5. Added some features to "harden" Exim a bit more against certain attacks:
4249 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4250 be put in Local/Makefile. This is like the never_users runtime option,
4251 but it cannot be overridden. The default setting is "root".
4253 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4254 prefix string with which any file named in a -C command line option
4257 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4258 is retained for -C and -D only if the caller of Exim is root. Without
4259 it, the exim user may also use -C and -D and retain privilege.
4261 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4262 command line option is disabled.
4264 6. Macro names set by the -D option must start with an upper case letter, just
4265 like macro names defined in the configuration file.
4267 7. Added "dereference=" facility to LDAP.
4269 8. Two instances of the typo "uknown" in the source files are fixed.
4271 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4272 the Configure-Makefile script screwed up while processing it.
4274 10. Incorporated PCRE 4.4.
4276 11. The SMTP synchronization check was not operating right at the start of an
4277 SMTP session. For example, it could not catch a HELO sent before the client
4278 waited for the greeting. There is now a check for outstanding input at the
4279 point when the greeting is written. Because of the duplex, asynchronous
4280 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4281 way, but not yet received, when the check is performed.
4283 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4284 on TCP/IP sockets, because this apparently causes some broken clients to
4287 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4288 unchanged) from the Cygwin maintainer.
4290 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4291 was supported instead of testing for mbx. Effectively a typo.
4293 15. The spa authenticator server code was not checking that the input it
4294 received was valid base64.
4296 16. The debug output line for the "set" modifier in ACLs was not showing the
4297 name of the variable that was being set.
4299 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4301 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4302 lookup on its own. Something else was needed to trigger the lookup. For
4303 example, a match in host_lookup or the need for a host name in a host list.
4304 Now, if $sender_host_name is referenced and the host name has not yet been
4305 looked up, a lookup is performed. If the lookup fails, the variable remains
4306 empty, and $host_lookup_failed is set to "1".
4308 19. Added "eqi" as a case-independent comparison operator.
4310 20. The saslauthd authentication condition could segfault if neither service
4311 nor realm was specified.
4313 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4314 error message that was logged was misleading, and incoming SMTP
4315 connections were dropped. The message is now more accurate, and temporary
4316 errors are given to SMTP connections.
4318 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4319 (except RSET) with a 421 error, until QUIT is received. However, it was
4320 failing to send a response to QUIT.
4322 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4323 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4324 ACL is now run after the helo_try_verify_hosts code.
4326 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4327 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4328 case-independent, so other case variants are also recognized). Apparently
4329 some systems use these upper case variants.
4331 25. If more than two messages were waiting for the same host, and a transport
4332 filter was specified for the transport, Exim sent two messages over the
4333 same TCP/IP connection, and then failed with "socket operation on non-
4334 socket" when it tried to send the third.
4336 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4338 27. The extern definition of crypt16() in expand.c was not being excluded when
4339 the OS had its own crypt16() function.
4341 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4342 as a preferred synonym for return_size_limit, both as an option and as an
4345 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4347 30. Changed the default configuration ACL to relax the local part checking rule
4348 for addresses that are not in any local domains. For these addresses,
4349 slashes and pipe symbols are allowed within local parts, but the sequence
4350 /../ is explicitly forbidden.
4352 31. SPA server authentication was not clearing the challenge buffer before
4355 32. log_message in a "warn" ACL statement was writing to the reject log as
4356 well as to the main log, which contradicts the documentation and doesn't
4357 seem right (because no rejection is happening). So I have stopped it.
4359 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4360 However, I am unable to do any testing of this.
4362 34. Fixed an infelicity in the appendfile transport. When checking directories
4363 for a mailbox, to see if any needed to be created, it was accidentally
4364 using path names with one or more superfluous leading slashes; tracing
4365 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4367 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4368 contents are added to the log line that is written for every discarded
4369 recipient. (Previously a log_message setting was ignored.)
4371 36. The ${quote: operator now quotes the string if it is empty.
4373 37. The install script runs exim in order to find its version number. If for
4374 some reason other than non-existence or emptiness, which it checks, it
4375 could not run './exim', it was installing it with an empty version number,
4376 i.e. as "exim-". This error state is now caught, and the installation is
4379 38. An argument was missing from the function that creates an error message
4380 when Exim fails to connect to the socket for saslauthd authentication.
4381 This could cause Exim to crash, or give a corrupted message.
4383 39. Added isip, isip4, and isip6 to ${if conditions.
4385 40. The ACL variables $acl_xx are now saved with the message, and can be
4386 accessed later in routers, transports, and filters.
4388 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4389 strings in the file are not string-expanded.
4391 42. If a MAIL command specified a SIZE value that was too large to fit into an
4392 int variable, the check against message_size_limit failed. Such values are
4393 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4394 day this will have to be increased, but I don't think I want to be around
4395 when emails are that large.
4402 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4403 iconv() is not standard in FreeBSD.
4405 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4406 IPv6 enabled. The observed symptom was a segmentation fault on return from
4407 the function os_common_find_running_interfaces() in src/os.c.
4409 3. In the check_special_case() function in daemon.c I had used "errno" as an
4410 argument name, which causes warnings on some systems. This was basically a
4411 typo, since it was named "eno" in the comments!
4413 4. The code that waits for the clock to tick (at a resolution of some fraction
4414 of a second) so as to ensure message-id uniqueness was always waiting for
4415 at least one whole tick, when it could have waited for less. [This is
4416 almost certainly not relevant at current processor speeds, where it is
4417 unlikely to ever wait at all. But we try to future-proof.]
4419 5. The function that sleeps for a time interval that includes fractions of a
4420 second contained a race. It did not block SIGALRM between setting the
4421 timer, and suspending (a couple of lines later). If the interval was short
4422 and the sigsuspend() was delayed until after it had expired, the suspension
4423 never ended. On busy systems this could lead to processes getting stuck for
4426 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4427 process, before it forks any delivery processes. The open lookup caching
4428 mechanism meant that the open file or database connection was passed into
4429 the delivery process. The problem was that delivery processes always tidy
4430 up cached lookup data. This could cause a problem for the next delivery
4431 process started by the queue runner, because the external queue runner
4432 process does not know about the closure. So the next delivery process
4433 still has data in the lookup cache. In the case of a file lookup, there was
4434 no problem because closing a file descriptor in a subprocess doesn't affect
4435 the parent. However, if the lookup was caching a connection to a database,
4436 the connection was closed, and the second delivery process was likely to
4437 see errors such as "PGSQL: query failed: server closed the connection
4438 unexpectedly". The problem has been fixed by closing all cached lookups
4439 in a queue runner before running a delivery process.
4441 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4442 seem to have the "const" qualifier which it has on other OS. I've
4445 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4446 *of the same type* (client or server) with the same public name that an
4447 error should be diagnosed.
4449 9. When Exim looked up a host name for an IP address, but failed to find the
4450 original IP address when looking up the host name (a safety check), it
4451 output the message "<ip address> does not match any IP for NULL", which was
4452 confusing, to say the least. The bug was that the host name should have
4453 appeared instead of "NULL".
4455 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4456 user that is built into the binary, and the -C or -D options is used, root
4457 privilege is dropped before the configuration file is read. In addition,
4458 logging is switched to stderr instead of the normal log files. If the
4459 configuration then re-defines the Exim user, the unprivileged environment
4460 is probably not what is expected, so Exim logs a panic warning message (but
4463 However, if deliver_drop_privilege is set, the unprivileged state may well
4464 be exactly what is intended, so the warning has been cut out in that case,
4465 and Exim is allowed to try to write to its normal log files.
4471 1. smtp_return_error_details was not giving details for temporary sender
4472 or receiver verification errors.
4474 2. Diagnose a configuration error if two authenticators have the same public
4477 3. Exim used not to create the message log file for a message until the first
4478 delivery attempt. This could be confusing when incoming messages were held
4479 for policy or load reasons. The message log file is now created at the time
4480 the message is received, and an initial "Received" line is written to it.
4482 4. The automatically generated man page for command line options had a minor
4483 bug that caused no ill effects; however, a more serious problem was that
4484 the procedure for building the man page automatically didn't always
4485 operate. Consequently, release 4.20 contains an out-of-date version. This
4486 shouldn't happen again.
4488 5. When building Exim with embedded Perl support, the script that builds the
4489 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4490 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4492 6. The freeze_tell option was not being used for messages that were frozen on
4493 arrival, either by an ACL or by local_scan().
4495 7. Added the smtp_incomplete_transaction log selector.
4497 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4498 was accepting AUTH without a new EHLO.
4500 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4501 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4504 10. Logging of TCP/IP connections (when configured) now happens in the main
4505 daemon process instead of the child process, so that the TCP/IP connection
4506 count is more accurate (but it can never be perfect).
4508 11. The use of "drop" in a nested ACL was not being handled correctly in the
4509 outer ACL. Now, if condition failure induced by the nested "drop" causes
4510 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4511 or "require"), the connection is dropped.
4513 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4514 that yield "discard" can now be used with an "accept" or a "discard" verb,
4515 but an error is generated for any others (because I can't see a useful way
4516 to define what should happen).
4518 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4519 are now processed in the same way as lines in the Exim configuration file.
4520 In particular, continuation lines are supported.
4522 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4524 15. Added -ti meaning -t -i.
4526 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4527 domains, and warn by logging if others are found.
4529 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4530 returned from GIFCONF when Exim is trying to find the list of interfaces on
4531 a host. The code in os.c has been modified to copy each ifreq to an aligned
4532 structure in all cases.
4534 Also, in some cases, the returned ifreq's were being copied to a 'struct
4535 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4536 means the last couple of bytes of an IPv6 address could be chopped if the
4537 ifreq contained only a normal sockaddr (14 bytes storage).
4539 18. Named domain lists were not supported in the hosts_treat_as_local option.
4540 An entry such as +xxxx was not recognized, and was treated as a literal
4543 19. Ensure that header lines added by a DATA ACL are included in the reject log
4544 if the ACL subsequently rejects the message.
4546 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4547 MD5 (which is deprecated).
4549 21. When testing a filter file using -bf, Exim was writing a message when it
4550 took the sender from a "From " line in the message, but it was not doing so
4551 when it took $return_path from a Return-Path: header line. It now does.
4553 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4554 with a valid header line field name (a series of printing characters
4555 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4557 23. Changed "disc" in the source to "disk" to conform to the documentation and
4558 the book and for uniformity.
4560 24. Ignore Sendmail's -Ooption=value command line item.
4562 25. When execve() failed while trying to run a command in a pipe transport,
4563 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4564 could be confused with a return value of 69 from the command itself. This
4565 has been changed to 127, the value the shell returns if it is asked to run
4566 a non-existent command. The wording for the related log line suggests a
4567 non-existent command as the problem.
4569 26. If received_header_text expands to an empty string, do not add a Received:
4570 header line to the message. (Well, it adds a token one on the spool, but
4571 marks it "old" so that it doesn't get used or transmitted.)
4573 27. Installed eximstats 1.28 (addition of -nt option).
4575 28. There was no check for failure on the call to getsockname() in the daemon
4576 code. This can fail if there is a shortage of resources on the system, with
4577 ENOMEM, for example. A temporary error is now given on failure.
4579 29. Contrary to the C standard, it seems that in some environments, the
4580 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4581 program. Exim now does this explicitly; it affects the formatting of
4582 timestamps using strftime().
4584 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4585 complaints. I've now initialized all the variables, to avoid this.
4587 32. Header lines added by a system filter were not being "seen" during
4588 transport-time rewrites.
4590 33. The info_callback() function passed to OpenSSL is set up with type void
4591 (*)(SSL *, int, int), as described somewhere. However, when calling the
4592 function (actually a macro) that sets it up, the type void(*)() is
4593 expected. I've put in a cast to prevent warnings from picky compilers.
4595 34. If a DNS black list lookup found a CNAME record, but there were no A
4596 records associated with the domain it pointed at, Exim crashed.
4598 35. If a DNS black list lookup returned more than one A record, Exim ignored
4599 all but the first. It now scans all returned addresses if a particular IP
4600 value is being sought. In this situation, the contents of the
4601 $dnslist_value variable are a list of all the addresses, separated by a
4604 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4605 to accept a host name and all its aliases if the forward lookup for any of
4606 them yielded the IP address of the incoming connection. Now it accepts only
4607 those names whose forward lookup yields the correct IP address. Any other
4608 names are discarded. This closes a loophole whereby a rogue DNS
4609 administrator could create reverse DNS records to break through a
4610 wildcarded host restriction in an ACL.
4612 37. If a user filter or a system filter that ran in a subprocess used any of
4613 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4614 the wrong values were passed to the pipe command ($thisaddress had the
4615 value of $0, $0 had the value of $1, etc). This bug was introduced by
4616 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4618 38. Improved the line breaking for long SMTP error messages from ACLs.
4619 Previously, if there was no break point between 40 and 75 characters, Exim
4620 left the rest of the message alone. Two changes have been made: (a) I've
4621 reduced the minimum length to 35 characters; (b) if it can't find a break
4622 point between 35 and 75 characters, it looks ahead and uses the first one
4623 that it finds. This may give the occasional overlong line, but at least the
4624 remaining text gets split now.
4626 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4627 file descriptors might be low, and that setting 1000 would always raise it.
4628 It turns out that in some environments, the limit is already over 1000 and
4629 that lowering it causes trouble. So now Exim takes care not to decrease it.
4631 40. When delivering a message, the value of $return_path is set to $sender_
4632 address at the start of routing (routers may change the value). By an
4633 oversight, this default was not being set up when an address was tested by
4634 -bt or -bv, which affected the outcome if any router or filter referred to
4637 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4638 log only when "message" or "log_message" are set. However, if one of the
4639 conditions was an address verification, or a call to a nested ACL, the
4640 messages generated by the underlying test were being passed through. This
4641 no longer happens. The underlying message is available in $acl_verify_
4642 message for both "message" and "log_message" expansions, so it can be
4643 passed through if needed.
4645 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4646 new expansion $bh_ to give the encoded byte string without charset
4647 translation. Translation happens only if iconv() is available; HAVE_ICONV
4648 indicates this at build time. HEADERS_CHARSET gives the charset to
4649 translate to; headers_charset can change it in the configuration, and
4650 "headers charset" can change it in an individual filter file.
4652 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4653 that creates RFC 2047 encoded "words" labels them as that charset instead
4654 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4655 expansion operator; (ii) when Exim creates a From: line for a local
4656 message; (iii) when a header line is rewritten to include a "phrase" part.
4658 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4659 buggy, causing it to skip the first lines of messages whose message ID
4660 ended in 'D'. This would not have bitten before Exim release 4.14, because
4661 message IDs were unlikely to end in 'D' before then. The effect was to have
4662 incorrect size information for certain domains.
4664 45. #include "config.h" was missing at the start of the crypt16.c module. This
4665 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4668 46. If there was a timeout during a "random" callout check, Exim treated it as
4669 a failure of the random address, and carried on sending RSET and the real
4670 address. If the delay was just some slowness somewhere, the response to the
4671 original RCPT would be taken as a response to RSET and so on, causing
4672 mayhem of various kinds.
4674 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4675 when I implemented it. It didn't allow for the fact that some option values
4676 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4677 the right test for positive values.
4679 48. Domain names in DNS records are case-independent. Exim always looks them up
4680 in lower case. Some resolvers return domain names in exactly the case they
4681 appear in the zone file, that is, they may contain uppercase letters. Not
4682 all resolvers do this - some return always lower case. Exim was treating a
4683 change of case by a resolver as a change of domain, similar to a widening
4684 of a domain abbreviation. This triggered its re-routing code and so it was
4685 trying to route what was effectively the same domain again. This normally
4686 caused routing to fail (because the router wouldn't handle the domain
4687 twice). Now Exim checks for this case specially, and just changes the
4688 casing of the domain that it ultimately uses when it transmits the message
4691 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4694 50. If a filter generated a file delivery with a non-absolute name (possible if
4695 no home directory exists for the router), the forbid_file option was not
4698 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4699 the existing equality matching.
4701 52. Exim was using ints instead of ino_t variables in some places where it was
4702 dealing with inode numbers.
4704 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4705 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4706 called TMPDIR, and if it finds it is different, it changes its value.
4708 54. The smtp_printf() function is now made available to local_scan() so
4709 additional output lines can be written before returning. There is also an
4710 smtp_fflush() function to enable the detection of a dropped connection.
4711 The variables smtp_input and smtp_batched_input are exported to
4714 55. Changed the default runtime configuration: the message "Unknown user"
4715 has been removed from the ACL, and instead placed on the localuser router,
4716 using the cannot_route_message feature. This means that any verification
4717 failures that generate their own messages won't get overridden. Similarly,
4718 the "Unrouteable address" message that was in the ACL for unverifiable
4719 relay addresses has also been removed.
4721 56. Added hosts_avoid_esmtp to the smtp transport.
4723 57. The exicyclog script was not checking for the esoteric option
4724 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4725 will work only if exicyclog is run under the appropriate euid.
4727 58. Following a discussion on the list, the rules by which Exim recognises line
4728 endings on incoming messages have been changed. The -dropcr and drop_cr
4729 options are now no-ops, retained only for backwards compatibility. The
4730 following line terminators are recognized: LF CRLF CR. However, special
4731 processing applies to CR:
4733 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4734 nor a local message in the state where . is a terminator.
4736 (ii) If a bare CR is encountered in a header line, an extra space is added
4737 after the line terminator so as not to end the header. The reasoning
4738 behind this is that bare CRs in header lines are most likely either
4739 to be mistakes, or people trying to play silly games.
4741 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4742 was being incorrectly given as 18 bytes larger than it should have been.
4743 This is a VOB (very old bug).
4745 60. This may never have affected anything current, but just in case it has:
4746 When the local host is found other than at the start of a list of hosts,
4747 the local host, those with the same MX, and any that follow, are discarded.
4748 When the list in question was part of a longer list of hosts, the following
4749 hosts (not currently being processed) were also being discarded. This no
4750 longer happens. I'm not sure if this situation could ever has previously
4753 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4756 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4757 header. This is something that is recommended in a new Internet Draft, and
4758 is something that is documented as being done by Sendmail. There are two
4759 possible values. For messages generated by the autoreply transport, Exim
4762 Auto-Submitted: auto-replied
4764 whereas for all other generated messages (e.g. bounces) it adds
4766 Auto-Submitted: auto-generated
4768 63. The "personal" condition in filters now includes a test for the
4769 Auto-Submitted: header. If it contains the string "auto-" the message it
4770 not considered personal.
4772 64. Added rcpt_include_affixes as a generic transport option.
4774 65. Added queue_only_override (default true).
4776 66. Added the syslog_duplication option.
4778 67. If what should have been the first header line of a message consisted of
4779 a space followed by a colon, Exim was mis-interpreting it as a header line.
4780 It isn't of course - it is syntactically invalid and should therefore be
4781 treated as the start of the message body. The misbehaviour could have
4782 caused a number of strange effects, including loss of data in subsequent
4783 header lines, and spool format errors.
4785 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4786 client host had authenticated. This control can now be exercised by an ACL
4787 for more flexibility.
4789 69. By default, callouts do not happen when testing with -bh. There is now a
4790 variant, -bhc, which does actually run the callout code, including
4791 consulting and updating the callout cache.
4793 70. Added support for saslauthd authentication, courtesy of Alexander
4796 71. If statvfs() failed on the spool or log directories while checking their
4797 size for availability, Exim confusingly gave the error "space shortage".
4798 Furthermore, in debugging mode it crashed with a floating point exception.
4799 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4800 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4801 serious problem, Exim now writes to the main and panic logs when this
4802 happens, with details of the failure. It then refuses to accept the
4803 incoming message, giving the message "spool directory problem" or "log
4804 directory problem" with a 421 code for SMTP messages.
4806 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4807 0, 1, and 2 exist, because some OS complain for execs without them (see
4808 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4809 descriptors. However, the code omitted to check that the open succeeded,
4810 causing mysterious errors if for some reason the permissions on /dev/null
4811 got screwed. Now Exim writes a message to the main and panic logs, and
4812 bombs out if it can't open /dev/null.
4814 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4815 interact so that it is all more flexible. It is supposed to remain
4816 backwards compatible. Also added extra_local_interfaces.
4818 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4819 to bomb out with an assertion failure - to the client this appears as a
4820 connection drop. This problem occurs in the part of the code that was taken
4821 from the Samba project. Fortunately, the assertion is in a very simple
4822 function, so I have fixed this by reproducing the function inline in the
4823 one place where it is called, and arranging for authentication to fail
4824 instead of killing the process with assert().
4826 75. The SPA client code was not working when the server requested OEM rather
4827 than Unicode encoding.
4829 76. Added code to make require_files with a specific uid setting more usable in
4830 the case where statting the file as root fails - usually a non-root-mounted
4831 NFS file system. When this happens and the failure is EACCES, Exim now
4832 forks a subprocess and does the per-uid checking as the relevant uid.
4834 77. Added process_log_path.
4836 78. If log_file_path was not explicitly set, a setting of check_log_space or
4837 check_log_inodes was ignored.
4839 79. If a space check for the spool or log partitions fails, the incident is now
4840 logged. Of course, in the latter case the data may get lost...
4842 80. Added the %p formatting code to string_format() so that it can be used to
4843 print addresses in debug_print(). Adjusted all the address printing in the
4844 debugging in store.c to use %p rather than %d.
4846 81. There was a concern that a line of code in smtp_in.c could overflow a
4847 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4848 initially expressed, the concern was not well-founded, because trailing
4849 spaces are removed early. However, if the trailing spaces were followed by
4850 a NULL, they did not get removed, so the overflow was possible. Two fixes
4853 (a) I re-wrote the offending code in a cleaner fashion.
4854 (b) If an incoming SMTP command contains a NULL character, it is rejected
4857 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4858 runs initgroups(), so that if the Exim user is in any additional groups,
4859 they will be used during message reception.
4865 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob