1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.550 2008/07/28 18:46:44 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added patch by Johannes Berg that expands the main option
10 "spamd_servers" if it starts with a dollar sign.
12 TK/02 Write list of recipients to X-Envelope-Sender header when building
13 the mbox-format spool file for content scanning (suggested by Jakob
16 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
17 (http://www.dcc-servers.net/) support via dccifd. Activated by
18 setting EXPERIMENTAL_DCC=yes in Local/Makefile. Check out
19 experimental_spec.txt for more documentation.
21 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
22 by Mark Daniel Reidel <mr@df.eu>.
24 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
25 When building exim an external PCRE library is now needed -
26 PCRE is a system library on the majority of modern systems.
27 See entry on PCRE_LIBS in EDITME file.
29 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
30 conversation. Added nologin parameter to request.
31 Patch contributed by Kirill Miazine
33 TF/01 Do not log submission mode rewrites if they do not change the address.
35 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
37 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
38 log files in place. Contributed by Roberto Lima
40 NM/04 Bugzilla 667: close socket used by dovecot authenticator
42 TF/03 Bugzilla 615: When checking the local_parts router precondition
43 after a local_part_suffix or local_part_prefix option, Exim now
44 does not use the address's named list lookup cache, since this
45 contains cached lookups for the whole local part.
47 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
48 Robert Millan. Documentation is in experimental-spec.txt
50 TF/04 Bugzilla 668: Fix parallel build (make -j).
52 NM/05 Bugzilla 437: Prevent Maildix aux files being created with mode 000
54 NM/05 Bugzilla 598: Improvedment to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
67 TK/01 Add preliminary DKIM support. Currently requires a forked version of
68 ALT-N's libdkim that I have put here:
69 http://duncanthrax.net/exim-experimental/
71 Note to Michael Haardt: I had to rename some vars in sieve.c. They
72 were called 'true' and it seems that C99 defines that as a reserved
73 keyword to be used with 'bool' variable types. That means you could
74 not include C99-style headers which use bools without triggering
75 build errors in sieve.c.
77 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
78 as mailq or other aliases. Changed the --help handling significantly
79 to do whats expected. exim_usage() emits usage/help information.
81 SC/01 Added the -bylocaldomain option to eximstats.
83 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
85 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
87 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
93 PH/01 Another patch from the Sieve maintainer.
95 PH/02 When an IPv6 address is converted to a string for single-key lookup
96 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
97 dots are used instead of colons so that keys in lsearch files need not
98 contain colons. This was done some time before quoting was made available
99 in lsearch files. However, iplsearch files do require colons in IPv6 keys
100 (notated using the quote facility) so as to distinguish them from IPv4
101 keys. This meant that lookups for IP addresses in host lists did not work
102 for iplsearch lookups.
104 This has been fixed by arranging for IPv6 addresses to be expressed with
105 colons if the lookup type is iplsearch. This is not incompatible, because
106 previously such lookups could never work.
108 The situation is now rather anomolous, since one *can* have colons in
109 ordinary lsearch keys. However, making the change in all cases is
110 incompatible and would probably break a number of configurations.
112 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
115 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
116 conversion specification without a maximum field width, thereby enabling
117 a rogue spamd server to cause a buffer overflow. While nobody in their
118 right mind would setup Exim to query an untrusted spamd server, an
119 attacker that gains access to a server running spamd could potentially
120 exploit this vulnerability to run arbitrary code as the Exim user.
122 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
123 $primary_hostname instead of what libspf2 thinks the hosts name is.
125 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
126 a directory entry by the name of the lookup key. Previously, if a
127 symlink pointed to a non-existing file or a file in a directory that
128 Exim lacked permissions to read, a lookup for a key matching that
129 symlink would fail. Now it is enough that a matching directory entry
130 exists, symlink or not. (Bugzilla 503.)
132 PH/03 The body_linecount and body_zerocount variables are now exported in the
135 PH/04 Added the $dnslist_matched variable.
137 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
138 This means they are set thereafter only if the connection becomes
141 PH/06 Added the client_condition to authenticators so that some can be skipped
142 by clients under certain conditions.
144 PH/07 The error message for a badly-placed control=no_multiline_responses left
145 "_responses" off the end of the name.
147 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
149 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
150 (without spaces) instead of just copying the configuration text.
152 PH/10 Added the /noupdate option to the ratelimit ACL condition.
154 PH/11 Added $max_received_linelength.
156 PH/12 Added +ignore_defer and +include_defer to host lists.
158 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
159 way in which PCRE > 7.0 is built.
161 PH/14 Implemented queue_only_load_latch.
163 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
164 MAIL command. The effect was to mangle the value on 64-bit systems.
166 PH/16 Another patch from the Sieve maintainer.
168 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
170 PH/18 If a system quota error occurred while trying to create the file for
171 a maildir delivery, the message "Mailbox is full" was not appended to the
172 bounce if the delivery eventually timed out. Change 4.67/27 below applied
173 only to a quota excession during the actual writing of the file.
175 PH/19 It seems that peer DN values may contain newlines (and other non-printing
176 characters?) which causes problems in log lines. The DN values are now
177 passed through string_printing() before being added to log lines.
179 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
180 and InterBase are left for another time.)
182 PH/21 Added message_body_newlines option.
184 PH/22 Guard against possible overflow in moan_check_errorcopy().
186 PH/23 POSIX allows open() to be a macro; guard against that.
188 PH/24 If the recipient of an error message contained an @ in the local part
189 (suitably quoted, of course), incorrect values were put in $domain and
190 $local_part during the evaluation of errors_copy.
196 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
197 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
200 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
201 issue a MAIL command.
203 PH/02 In an ACL statement such as
205 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
207 if a client was not listed at all, or was listed with a value other than
208 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
209 the condition was not true (as it should be), so access was not denied.
210 The bug was that the ! inversion was incorrectly passed on to the second
211 item. This has been fixed.
213 PH/03 Added additional dnslists conditions == and =& which are different from
214 = and & when the dns lookup returns more than one IP address.
216 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
217 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
219 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
220 FSYNC, which compiles an option called disable_fsync that allows for
221 bypassing fsync(). The documentation is heavily laced with warnings.
223 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
225 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
226 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
227 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
228 including adding "make clean"; (3) Added -fPIC when compiling the test
229 dynamically loaded module, to get rid of a warning.
231 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
232 message fails, move_frozen_messages = true and ignore_bounce_errors_after
233 = 0s. The bug is otherwise harmless.
235 PH/07 There was a bug in the dovecot authenticator such that the value of
236 $auth1 could be overwritten, and so not correctly preserved, after a
237 successful authentication. This usually meant that the value preserved by
238 the server_setid option was incorrect.
240 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
242 PH/09 Installed PCRE release 7.0.
244 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
245 run for batched SMTP input. It is now run at the start of every message
246 in the batch. While fixing this I discovered that the process information
247 (output by running exiwhat) was not always getting set for -bs and -bS
248 input. This is fixed, and it now also says "batched" for BSMTP.
250 PH/11 Added control=no_pipelining.
252 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
253 patch, slightly modified), and move the expansion of helo_data till after
254 the connection is made in the smtp transport (so it can use these
257 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
259 PH/14 Added log_selector = +pid.
261 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
263 PH/16 Add ${if forany and ${if forall.
265 PH/17 Added dsn_from option to vary the From: line in DSNs.
267 PH/18 Flush SMTP output before performing a callout, unless control =
268 no_callout_flush is set.
270 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
271 was true (the default) a successful delivery failed to delete the retry
272 item, thus causing premature timeout of the address. The bug is now
275 PH/20 Added hosts_avoid_pipelining to the smtp transport.
277 PH/21 Long custom messages for fakedefer and fakereject are now split up
278 into multiline reponses in the same way that messages for "deny" and
279 other ACL rejections are.
281 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
282 with slight modification.
284 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
285 draft, changing the syntax and factoring some duplicate code".
287 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
288 for deliveries of the second and subsequent messages over the same SMTP
291 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
292 ${reduce, with only minor "tidies".
294 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
296 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
297 expansion side effects.
299 PH/27 When a message times out after an over-quota error from an Exim-imposed
300 quota, the bounce message says "mailbox is full". This message was not
301 being given when it was a system quota that was exceeded. It now should
304 MH/03 Made $recipients available in local_scan(). local_scan() already has
305 better access to the recipient list through recipients_list[], but
306 $recipients can be useful in postmaster-provided expansion strings.
308 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
309 in the case of a MAIL command with additional options following the
310 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
311 were accidentally chopped off.
313 PH/29 SMTP synchronization checks are implemented when a command is read -
314 there is a check that no more input is waiting when there shouldn't be
315 any. However, for some commands, a delay in an ACL can mean that it is
316 some time before the response is written. In this time, more input might
317 arrive, invalidly. So now there are extra checks after an ACL has run for
318 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
319 pipelining has not been advertised.
321 PH/30 MH's patch to allow iscntrl() characters to be list separators.
323 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
324 returned in the SMTP response when smtp_return_error_details was false.
327 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
328 instead of the C I/O that was originally supplied, because problems were
331 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
332 Exim which did not show up earlier: it was assuming that a call to
333 SSL_CTX_set_info_callback() might give an error value. In fact, there is
334 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
335 was a macro that became an assignment, so it seemed to work. This has
336 changed to a proper function call with a void return, hence the compile
337 error. Exim's code has been fixed.
339 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
342 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
344 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
346 PH/37 If a message is not accepted after it has had an id assigned (e.g.
347 because it turns out to be too big or there is a timeout) there is no
348 "Completed" line in the log. When some messages of this type were
349 selected by exigrep, they were listed as "not completed". Others were
350 picked up by some special patterns. I have improved the selection
351 criteria to be more general.
353 PH/38 The host_find_failed option in the manualroute router can now be set
354 to "ignore", to completely ignore a host whose IP address cannot be
355 found. If all hosts are ignored, the behaviour is controlled by the new
356 host_all_ignored option.
358 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
359 homing or because of multiple MX records with /mx) generated more than
360 one IP address, and the following item turned out to be the local host,
361 all the secondary addresses of the first item were incorrectly removed
362 from the list, along with the local host and any following hosts (which
363 is what is supposed to happen).
365 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
366 whoever called Exim into the -H file. In the case of the daemon it was
367 behaving confusingly. When first started, it used values for whoever
368 started the daemon, but after a SIGHUP it used the Exim user (because it
369 calls itself on a restart). I have changed the code so that it now always
372 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
373 message are rejected with the same error (e.g. no authentication or bad
374 sender address), and a DATA command is nevertheless sent (as can happen
375 with PIPELINING or a stupid MUA), the error message that was given to the
376 RCPT commands is included in the rejection of the DATA command. This is
377 intended to be helpful for MUAs that show only the final error to their
380 PH/42 Another patch from the Sieve maintainer.
382 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
383 Eximstats - Fixed some broken HTML links and added missing column headers
385 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
386 columns for Rejects, Temp Rejects, Ham, and Spam rows.
388 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
390 PH/43 Yet another patch from the Sieve maintainer.
392 PH/44 I found a way to check for a TCP/IP connection going away before sending
393 the response to the final '.' that terminates a message, but only in the
394 case where the client has not sent further data following the '.'
395 (unfortunately, this is allowed). However, in many cases there won't be
396 any further data because there won't be any more messages to send. A call
397 to select() can be used: if it shows that the input is "ready", there is
398 either input waiting, or the socket has been closed. An attempt to read
399 the next input character can distinguish the two cases. Previously, Exim
400 would have sent an OK response which the client would never have see.
401 This could lead to message repetition. This fix should cure that, at
402 least in a lot of common cases.
404 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
405 advertised in response to EHLO.
411 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
412 fixed by 4.65/MH/01 (is this a record?) are fixed:
414 (i) An empty string was always treated as zero by the numeric comparison
415 operators. This behaviour has been restored.
417 (ii) It is documented that the numeric comparison operators always treat
418 their arguments as decimal numbers. This was broken in that numbers
419 starting with 0 were being interpreted as octal.
421 While fixing these problems I realized that there was another issue that
422 hadn't been noticed. Values of message_size_limit (both the global option
423 and the transport option) were treated as octal if they started with 0.
424 The documentation was vague. These values are now always treated as
425 decimal, and I will make that clear in the documentation.
431 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
432 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
435 MH/01 Don't check that the operands of numeric comparison operators are
436 integers when their expansion is in "skipping" mode (fixes bug
437 introduced by 4.64-PH/07).
439 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
440 child addresses, Exim now panics and dies. Previously, because the count
441 is held in a short int, deliveries were likely to be lost. As such a
442 large number of recipients for a single message is ridiculous
443 (performance will be very, very poor), I have chosen to impose a limit
444 rather than extend the field.
450 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
451 leftover -K file (the existence of which was triggered by #402).
452 While we were at it, introduced process PID as part of the -K
453 filename. This should rule out race conditions when creating
456 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
457 processing considerably. Previous code took too long for large mails,
458 triggering a timeout which in turn triggers #401.
460 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
461 in the DK code in transports.c. sendfile() is not really portable,
462 hence the _LINUX specificness.
464 TF/01 In the add_headers option to the mail command in an Exim filter,
465 there was a bug that Exim would claim a syntax error in any
466 header after the first one which had an odd number of characters
469 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
470 callout verification, Exim cached a "reject" for the entire domain. This
471 is correct for most verifications, but it is not correct for a recipient
472 verification with use_sender or use_postmaster set, because in that case
473 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
474 case of MAIL FROM:<> rejection from other early rejections (e.g.
475 rejection of HELO). When verifying a recipient using a non-null MAIL
476 address, the cache is ignored if it shows MAIL FROM:<> rejection.
477 Whatever the result of the callout, the value of the domain cache is
478 left unchanged (for any other kind of callout, getting as far as trying
479 RCPT means that the domain itself is ok).
481 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
484 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
485 manifest itself as EPIPE rather than ECONNECT. When tidying away a
486 session, the daemon ignores ECONNECT errors and logs others; it now
487 ignores EPIPE as well.
489 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
490 (quoted-printable decoding).
492 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
493 later the small subsequent patch to fix an introduced bug.
495 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
497 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
499 PH/08 An error is now given if message_size_limit is specified negative.
501 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
502 to be given (somewhat) arbitrary names.
504 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
507 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
508 miscellaneous code fixes
510 PH/10 Added the log_reject_target ACL modifier to specify where to log
513 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
514 hostname. This is wrong, because it relates to the incoming message (and
515 probably the interface on which it is arriving) and not to the outgoing
516 callout (which could be using a different interface). This has been
517 changed to use the value of the helo_data option from the smtp transport
518 instead - this is what is used when a message is actually being sent. If
519 there is no remote transport (possible with a router that sets up host
520 addresses), $smtp_active_hostname is used.
522 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
523 tweaks were necessary in order to get it to work (see also 21 below):
524 (a) The code assumed that strncpy() returns a negative number on buffer
525 overflow, which isn't the case. Replaced with Exim's string_format()
527 (b) There were several signed/unsigned issues. I just did the minimum
528 hacking in of casts. There is scope for a larger refactoring.
529 (c) The code used strcasecmp() which is not a standard C function.
530 Replaced with Exim's strcmpic() function.
531 (d) The code set only $1; it now sets $auth1 as well.
532 (e) A simple test gave the error "authentication client didn't specify
533 service in request". It would seem that Dovecot has changed its
534 interface. Fortunately there's a specification; I followed it and
535 changed what the client sends and it appears to be working now.
537 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
540 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
541 address A is aliased to B and C, where B exists and C does not. Without
542 -v the output is "A verified" because verification stops after a
543 successful redirection if more than one address is generated. However,
544 with -v the child addresses are also verified. Exim was outputting "A
545 failed to verify" and then showing the successful verification for C,
546 with its parentage. It now outputs "B failed to verify", showing B's
547 parentage before showing the successful verification of C.
549 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
550 look up a TXT record in a specific list after matching in a combined
553 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
554 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
555 they consult the DNS. I had assumed they would set it the way they
556 wanted; and indeed my experiments on Linux seem to show that in some
557 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
558 To be on the safe side, however, I have now made the interface to
559 host_find_byname() similar to host_find_bydns(), with an argument
560 containing the DNS resolver options. The host_find_byname() function now
561 sets these options at its start, just as host_find_bydns() does. The smtp
562 transport options dns_qualify_single and dns_search_parents are passed to
563 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
564 of host_find_byname() use the default settings of RES_DEFNAMES
565 (qualify_single) but not RES_DNSRCH (search_parents).
567 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
568 spool_read_header() do less string testing, by means of a preliminary
569 switch on the second character of optional "-foo" lines. (This is
570 overdue, caused by the large number of possibilities that now exist.
571 Originally there were few.) While I was there, I also converted the
572 str(n)cmp tests so they don't re-test the leading "-" and the first
573 character, in the hope this might squeeze out yet more improvement.
575 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
576 flag allowing group syntax was set by the header_syntax check but not
577 turned off, possible causing trouble later; (2) The flag was not being
578 set at all for the header_verify test, causing "group"-style headers to
579 be rejected. I have now set it in this case, and also caused header_
580 verify to ignore an empty address taken from a group. While doing this, I
581 came across some other cases where the code for allowing group syntax
582 while scanning a header line wasn't quite right (mostly, not resetting
583 the flag correctly in the right place). These bugs could have caused
584 trouble for malformed header lines. I hope it is now all correct.
586 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
587 with the "reply" argument non-NULL. The code, however (which originally
588 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
589 but it didn't always do it. This confused somebody who was copying the
590 code for some other use. I have removed all the tests.
592 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
593 feature that was used to support insecure browsers during the U.S. crypto
594 embargo. It requires special client support, and Exim is probably the
595 only MTA that supported it -- and would never use it because real RSA is
596 always available. This code has been removed, because it had the bad
597 effect of slowing Exim down by computing (never used) parameters for the
598 RSA_EXPORT functionality.
600 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
601 authenticator to fail if there's a tab character in the incoming data
602 (there should never be unless someone is messing about, as it's supposed
603 to be base64-encoded). Also added, on Timo's advice, the "secured" option
604 if the connection is using TLS or if the remote IP is the same as the
605 local IP, and the "valid-client-cert option" if a client certificate has
608 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
609 authenticators. This can be used for authorization after authentication
610 succeeds. (In the case of plaintext, it servers for both authentication
613 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
614 if any retry times were supplied.
616 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
617 connection, where there is no client IP address to check. In this
618 situation, the verify now always succeeds.
620 PH/25 Applied John Jetmore's -Mset patch.
622 PH/26 Added -bem to be like -Mset, but loading a message from a file.
624 PH/27 In a string expansion for a processed (not raw) header when multiple
625 headers of the same name were present, leading whitespace was being
626 removed from all of them, but trailing whitespace was being removed only
627 from the last one. Now trailing whitespace is removed from each header
628 before concatenation. Completely empty headers in a concatenation (as
631 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
632 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
634 PH/29 [Removed. This was a change that I later backed out, and forgot to
635 correct the ChangeLog entry (that I had efficiently created) before
636 committing the later change.]
638 PH/30 Exim was sometimes attempting to deliver messages that had suffered
639 address errors (4xx response to RCPT) over the same connection as other
640 messages routed to the same hosts. Such deliveries are always "forced",
641 so retry times are not inspected. This resulted in far too many retries
642 for the affected addresses. The effect occurred only when there were more
643 hosts than the hosts_max_try setting in the smtp transport when it had
644 the 4xx errors. Those hosts that it had tried were not added to the list
645 of hosts for which the message was waiting, so if all were tried, there
646 was no problem. Two fixes have been applied:
648 (i) If there are any address or message errors in an SMTP delivery, none
649 of the hosts (tried or untried) are now added to the list of hosts
650 for which the message is waiting, so the message should not be a
651 candidate for sending over the same connection that was used for a
652 successful delivery of some other message. This seems entirely
653 reasonable: after all the message is NOT "waiting for some host".
654 This is so "obvious" that I'm not sure why it wasn't done
655 previously. Hope I haven't missed anything, but it can't do any
656 harm, as the worst effect is to miss an optimization.
658 (ii) If, despite (i), such a delivery is accidentally attempted, the
659 routing retry time is respected, so at least it doesn't keep
660 hammering the server.
662 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
663 in ${readsocket because some servers need this prod.
665 PH/32 Added some extra debug output when updating a wait-xxx database.
667 PH/33 The hint "could be header name not terminated by colon", which has been
668 given for certain expansion errors for a long time, was not being given
669 for the ${if def:h_colon_omitted{... case.
671 PH/34 The spec says: "With one important exception, whenever a domain list is
672 being scanned, $domain contains the subject domain." There was at least
673 one case where this was not true.
675 PH/35 The error "getsockname() failed: connection reset by peer" was being
676 written to the panic log as well as the main log, but it isn't really
677 panic-worthy as it just means the connection died rather early on. I have
678 removed the panic log writing for the ECONNRESET error when getsockname()
681 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
682 runs only) independently of the message's sender address. This meant
683 that, if the 4xx error was in fact related to the sender, a different
684 message to the same recipient with a different sender could confuse
685 things. In particualar, this can happen when sending to a greylisting
686 server, but other circumstances could also provoke similar problems.
687 I have changed the default so that the retry time for these errors is now
688 based a combination of the sender and recipient addresses. This change
689 can be overridden by setting address_retry_include_sender=false in the
692 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
693 remote server are returned as part of bounce messages. This was not
694 happening for LMTP over a pipe (the lmtp transport), but now it is the
695 same for both kinds of LMTP.
697 PH/38 Despite being documented as not happening, Exim was rewriting addresses
698 in header lines that were in fact CNAMEs. This is no longer the case.
700 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
701 and queue runs started by the daemon processed all messages. This has
702 been fixed so that -R and -S can now usefully be given with -q<time>.
704 PH/40 Import PCRE release 6.7 (fixes some bugs).
706 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
708 PH/42 Give an error if -q is specified more than once.
710 PH/43 Renamed the variables $interface_address and $interface_port as
711 $received_ip_address and $received_port, to make it clear that these
712 values apply to message reception, and not to the outgoing interface when
713 a message is delivered. (The old names remain recognized, of course.)
715 PH/44 There was no timeout on the connect() call when using a Unix domain
716 socket in the ${readsocket expansion. There now is.
718 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
719 be meaningful with "accept".
721 SC/01 Eximstats V1.43
722 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
724 SC/02 Eximstats V1.44
725 Use a glob alias rather than an array ref in the generated
726 parser. This improves both readability and performance.
728 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
729 Collect SpamAssassin and rejection statistics.
730 Don't display local sender or destination tables unless
731 there is data to show.
732 Added average volumes into the top table text output.
734 SC/04 Eximstats V1.46
735 Collect data on the number of addresses (recipients)
736 as well as the number of messages.
738 SC/05 Eximstats V1.47
739 Added 'Message too big' to the list of mail rejection
740 reasons (thanks to Marco Gaiarin).
742 SC/06 Eximstats V1.48
743 Mainlog lines which have GMT offsets and are too short to
744 have a flag are now skipped.
746 SC/07 Eximstats V1.49 (Alain Williams)
747 Added the -emptyok flag.
749 SC/08 Eximstats V1.50
750 Fixes for obtaining the IP address from reject messages.
752 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
753 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
754 whitesspace changes from 4.64-PH/27
756 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
759 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
760 are found, allow negative numbers in numeric criteria)
762 JJ/06 exipick.20061117.2, added new $message_body_missing variable
764 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
765 to match changes made in 4.64-PH/43
767 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
769 PH/47 Put in an explicit test for a DNS lookup of an address record where the
770 "domain" is actually an IP address, and force a failure. This locks out
771 those revolvers/nameservers that support "A-for-A" lookups, in
772 contravention of the specifications.
774 PH/48 When a host name was looked up from an IP address, and the subsequent
775 forward lookup of the name timed out, the host name was left in
776 $sender_host_name, contrary to the specification.
778 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
779 restricted to single-key lookups, Exim was not diagnosing an error if
780 * or *@ was used with a query-style lookup.
782 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
784 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
785 long ago, but noone interested enough thought of it. Let's just say that
786 the "1.1" means that there are some new functions that weren't there at
787 some point in the past.
789 PH/51 Error processing for expansion failure of helo_data from an smtp
790 transport during callout processing was broken.
792 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
793 tested/used via the -bh/-bhc/-bs options.
795 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
796 bug, fixed in subsequent PCRE releases).
798 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
799 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
801 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
807 SC/01 Use a glob alias rather than an array ref in eximstats generated
808 parser. This improves both readability and performance.
810 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
811 Don't display local sender or destination tables in eximstats unless
812 there is data to show.
813 Added average volumes into the eximstats top table text output.
815 SC/03 Collect data on the number of addresses (recipients) as well
816 as the number of messages in eximstats.
818 TF/01 Correct an error in the documentation for the redirect router. Exim
819 does (usually) call initgroups() when daemonizing.
821 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
822 with consistent privilege compared to when running as a daemon.
824 TF/03 Note in the spec that $authenticated_id is not set for local
825 submissions from trusted users.
827 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
828 Thanks to Dean Brooks <dean@iglou.com> for the patch.
830 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
831 by adding some example configuration directives to the default
832 configuration file. A little bit of work is required to uncomment the
833 directives and define how usernames and passwords are checked, but
834 there is now a framework to start from.
836 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
837 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
838 without this. I don't know how relevant this is to other LDAP libraries.
840 PH/02 Add the verb name to the "unknown ACL verb" error.
842 PH/03 Magnus Holmgren's patch for filter_prepend_home.
844 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
846 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
847 directory not expanded when it should be if an expanded home directory
848 was set for the address (which is overridden by the transport).
850 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
853 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
854 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
855 because it is too late at that time, and has no effect.
857 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
858 security issue with \' (bugzilla #107). I could not use the
859 PQescapeStringConn() function, because it needs a PGconn value as one of
862 PH/08 When testing addresses using -bt, indicate those final addresses that
863 are duplicates that would not cause an additional delivery. At least one
864 person was confused, thinking that -bt output corresponded to deliveries.
865 (Suppressing duplicates isn't a good idea as you lose the information
866 about possibly different redirections that led to the duplicates.)
868 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
869 systems where poll() doesn't work, in particular OS X.
871 PH/10 Added more information to debugging output for retry time not reached.
873 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
874 operations in malware.c.
876 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
879 PH/13 If write_rejectlog was set false when logging was sent to syslog with
880 syslog_duplication set false, log lines that would normally be written
881 both the the main log and to the reject log were not written to syslog at
884 PH/14 In the default configuration, change the use of "message" in ACL warn
885 statements to "add_header".
887 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
888 not followed by a command (e.g. "seen endif").
890 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
891 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
894 PH/17 Added extra conditions to the default value of delay_warning_condition
898 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
899 { match{$h_precedence:}{(?i)bulk|list|junk} } \
900 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
903 The Auto-Submitted: and various List- headers are standardised, whereas I
904 don't think Precedence: ever was.
906 PH/18 Refactored debugging code in route_finduser() to show more information,
907 in particular, the error code if getpwnam() issues one.
909 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
910 This is apparently needed in addition to the PH/07 change above to avoid
911 any possible encoding problems.
913 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
914 but not after initializing Perl.
916 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
917 output them only if debugging. By default they are written stderr,
918 apparently, which is not desirable.
920 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
923 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
926 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
928 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
929 authenticated or an ident call has been made. Suppress the default
930 values for $authenticated_id and $authenticated_sender (but permit -oMai
931 and -oMas) when testing with -bh.
933 PH/24 Re-jigged the order of the tests in the default configuration so that the
934 tests for valid domains and recipients precede the DNS black list and CSA
935 tests, on the grounds that those ones are more expensive.
937 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
938 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
939 This bug exists in every version of Exim that I still have, right back to
942 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
943 However, an attempt to turn on case-sensitivity in a regex key by
944 including (?-i) didn't work because the subject string was already
945 lowercased, and the effects were non-intuitive. It turns out that a
946 one-line patch can be used to allow (?-i) to work as expected.
952 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
953 other effects) broke the use of negated acl sub-conditions.
955 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
958 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
959 "Deny" causes Exim to reject the incoming connection with a 554 error.
960 Unfortunately, if there is a major crisis, such as a disk failure,
961 tcp-wrappers gives "deny", whereas what one would like would be some
962 kind of temporary error. A kludge has been added to help with this.
963 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
964 554 error is used if errno is still zero or contains ENOENT (which occurs
965 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
968 PH/03 Add -lutil to the default FreeBSD LIBS setting.
970 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
971 errors. Otherwise a message that provokes a temporary error (when other
972 messages do not) can cause a whole host to time out.
974 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
975 addresses were routed directly to files or pipes from a redirect router.
976 File deliveries just didn't batch; pipe deliveries might have suffered
979 PH/06 A failure to get a lock for a hints database would erroneously always say
980 "Failed to get write lock", even when it was really a read lock.
982 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
983 of 0600. This has been changed to use the value of the lockfile_mode
984 option (which defaults to 0600).
986 PH/08 Applied small patch from the Sieve maintainer.
988 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
989 folder from quota calculations, a direct delivery into this folder messed
990 up the contents of the maildirsize file. This was because the regex was
991 used only to exclude .Trash (or whatever) when the size of the mailbox
992 was calculated. There was no check that a delivery was happening into an
993 excluded directory. This bug has been fixed by ignoring all quota
994 processing for deliveries into excluded directories.
996 PH/10 Added the maildirfolder_create_regex option to appendfile.
1002 PH/01 The code for finding all the local interface addresses on a FreeBSD
1003 system running IPv6 was broken. This may well have applied to all BSD
1004 systems, as well as to others that have similar system calls. The broken
1005 code found IPv4 interfaces correctly, but gave incorrect values for the
1006 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1007 that it would not match correctly against @[] and not recognize the IPv6
1010 PH/02 The ipliteral router was not recognizing addresses of the form user@
1011 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1013 PH/03 Added disable_ipv6.
1015 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1016 decoded form, because it is most often used to construct To: headers
1017 lines in autoreplies, and the decoded form may well be syntactically
1018 invalid. However, $reply_address has leading white space removed, and all
1019 newlines turned into spaces so that the autoreply transport does not
1022 PH/05 If group was specified without a user on a router, and no group or user
1023 was specified on a transport, the group from the router was ignored.
1025 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1026 for visible compile-time settings that can be used to change these
1027 numbers, for those that want even more. Backwards compatibility with old
1028 spool files has been maintained. However, going back to a previous Exim
1029 release will lost any variables that are in spool files.
1031 PH/07 Two small changes when running in the test harness: increase delay when
1032 passing a TCP/IP connection to a new process, in case the original
1033 process has to generate a bounce, and remove special handling of
1034 127.0.0.2 (sic), which is no longer necessary.
1036 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1037 be the same on different OS.
1039 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1042 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1043 whether --show-vars was specified or not
1045 JJ/02 exipick: Added support for new ACL variable spool format introduced
1048 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1049 syntactically invalid From: or Reply-to: line, and a filter used this to
1050 generate an autoreply, and therefore failed to obtain an address for the
1051 autoreply, Exim could try to deliver to a non-existent relative file
1052 name, causing unrelated and misleading errors. What now happens is that
1053 it logs this as a hard delivery error, but does not attempt to create a
1056 PH/11 The exinext utility has a -C option for testing purposes, but although
1057 the given file was scanned by exinext itself; it wasn't being passed on
1058 when Exim was called.
1060 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1061 an end-of-file indication when reading a command response.
1063 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1064 compiled. In many other places in Exim, IPv6 addresses are always
1065 recognized, so I have changed this. It also means that IPv4 domain
1066 literals of the form [IPV4:n.n.n.n] are now always recognized.
1068 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1069 used if the router is not running as root, for example, when verifying at
1070 ACL time, or when using -bh. The debugging output from this situation was
1071 non-existent - all you got was a failure to exec. I have made two
1074 (a) Failures to set uid/gid, the current directory, or a process leader
1075 in a subprocess such as that created by queryprogram now generate
1076 suitable debugging ouput when -d is set.
1078 (b) The queryprogram router detects when it is not running as root,
1079 outputs suitable debugging information if -d is set, and then runs
1080 the subprocess without attempting to change uid/gid.
1082 PH/15 Minor change to Makefile for building test_host (undocumented testing
1085 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1086 additional section of a DNS packet that returns MX or SRV records.
1087 Instead, it always explicitly searches for A/AAAA records. This avoids
1088 major problems that occur when a DNS server includes only records of one
1089 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1090 fixed another bug: if SRV records were looked up and the corresponding
1091 address records were *not* found in the additional section, the port
1092 values from the SRV records were lost.
1094 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1095 using the correct key (the original address) when searching the retry
1096 rules in order to find which one to use for generating the retry hint.
1098 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1099 adding the default one. Similarly, if it contains a Reply-To: header, the
1100 errors_reply_to option, if set, is not used.
1102 PH/19 When calculating a retry time, Exim used to measure the "time since
1103 failure" by looking at the "first failed" field in the retry record. Now
1104 it does not use this if it is later than than the arrival time of the
1105 message. Instead it uses the arrival time. This makes for better
1106 behaviour in cases where some deliveries succeed, thus re-setting the
1107 "first failed" field. An example is a quota failure for a huge message
1108 when small messages continue to be delivered. Without this change, the
1109 "time since failure" will always be short, possible causing more frequent
1110 delivery attempts for the huge message than are intended.
1111 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1113 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1114 $1, $2, $3) because the numerical variables can be reset during some
1115 expansion items (e.g. "match"), thereby losing the authentication data.
1117 PH/21 Make -bV show the size of off_t variables so that the test suite can
1118 decide whether to run tests for quotas > 2G.
1120 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1121 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1122 filecount value is greater than 2G or if a quota value is greater than 2G
1123 on a system where the size of off_t is not greater than 4, a panic error
1126 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1127 never match. The debug and -bh output now contains an explicit error
1128 message indicating a malformed IPv4 address or mask.
1130 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1131 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1132 PH/23 above applies.
1134 PH/25 Do not write to syslog when running in the test harness. The only
1135 occasion when this arises is a failure to open the main or panic logs
1136 (for which there is an explicit test).
1138 PH/26 Added the /no_tell option to "control=freeze".
1140 PH/27 If a host name lookup failed very early in a connection, for example, if
1141 the IP address matched host_lookup and the reverse lookup yielded a name
1142 that did not have a forward lookup, an error message of the form "no IP
1143 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1144 could be logged. Now it outputs the IP address instead of "NULL".
1146 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1147 allows the sender and the authenticated sender to be set when
1148 submitting a message from within Exim. Since child_open_exim() is
1149 documented for local_scan(), the new function should be too.
1151 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1152 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1153 results in an empty string is now treated as unset.
1155 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1157 PH/31 Added sender_verify_fail logging option.
1159 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1160 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1161 tidied the source and removed it altogether.
1163 PH/33 When a queue run was abandoned because the load average was too high, a
1164 log line was always written; now it is written only if the queue_run log
1165 selector is set. In addition, the log line for abandonment now contains
1166 information about the queue run such as the pid. This is always present
1167 in "start" and "stop" lines but was omitted from the "abandon" line.
1169 PH/34 Omit spaces between a header name and the colon in the error message that
1170 is given when verify = headers_syntax fails (if there are lots of them,
1171 the message gets confusing).
1173 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1174 names, as there are now some PTR records that contain slashes. This check
1175 is only to protect against broken name servers that fall over on strange
1176 characters, so the fact that it applies to all lookups doesn't matter.
1178 PH/36 Now that the new test suite is complete, we can remove some of the
1179 special code in Exim that was needed for the old test suite. For example,
1180 sorting DNS records because real resolvers return them in an arbitrary
1181 order. The new test suite's fake resolver always returns records in the
1184 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1185 bounces) except when queue_only is set, to avoid logging races between
1186 the different processes.
1188 PH/38 Panic-die if .include specifies a non-absolute path.
1190 PH/39 A tweak to the "H" retry rule from its user.
1192 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1193 a label. They prevented compilation on older perls.
1195 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1196 a warning to be raised on newish perls.
1198 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1199 on queue. Changes to match documented behaviour of showing count of
1200 messages matching specified criteria.
1202 PH/40 Changed the default ident timeout from 30s to 5s.
1204 PH/41 Added support for the use of login_cap features, on those BSD systems
1205 that have them, for controlling the resources used by pipe deliveries.
1207 PH/42 The content-scanning code uses fopen() to create files in which to put
1208 message data. Previously it was not paying any attention to the mode of
1209 the files. Exim runs with umask(0) because the rest of the code creates
1210 files with open(), and sets the required mode explicitly. Thus, these
1211 files were ending up world-writeable. This was not a big issue, because,
1212 being within the spool directory, they were not world-accessible. I have
1213 created a function called modefopen, which takes an additional mode
1214 argument. It sets umask(777), creates the file, chmods it to the required
1215 mode, then resets the umask. All the relevant calls to fopen() in the
1216 content scanning code have been changed to use this function.
1218 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1219 to 24 hours. This avoids potential overflow problems when processing G
1220 and H retry rules. I suspect nobody ever tinkers with this value.
1222 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1224 PH/45 When the plaintext authenticator is running as a client, the server's
1225 challenges are checked to ensure they are valid base64 strings. By
1226 default, the authentication attempt is cancelled if an invalid string is
1227 received. Setting client_ignore_invalid_base64 true ignores these errors.
1228 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1229 they are received. Thus, the responses can be made to depend on the
1230 challenges. If an invalid string is ignored, an empty string is placed in
1233 PH/46 Messages that are created by the autoreply transport now contains a
1234 References: header, in accordance with RFCs 2822 and 3834.
1236 PH/47 Added authenticated_sender_force to the smtp transport.
1238 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1240 PH/49 Installed latest patch from the Sieve maintainer.
1242 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1243 was also set, the appendfile transport was unnecessarily scanning a
1244 directory of message files (e.g. for maildir delivery) to find the count
1245 of files (along with the size), even though it did not need this
1246 information. It now does the scan only if it needs to find either the
1247 size of the count of files.
1249 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1251 PH/52 Two bugs concerned with error handling when the smtp transport is
1254 (i) Exim was not creating retry information for temporary errors given
1255 for individual recipients after the DATA command when the smtp transport
1256 was used in LMTP mode. This meant that they could be retried too
1257 frequently, and not timed out correctly.
1259 (ii) Exim was setting the flag that allows error details to be returned
1260 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1261 recipients that were returned after the DATA command.
1263 PH/53 This is related to PH/52, but is more general: for any failing address,
1264 when detailed error information was permitted to be returned to the
1265 sender, but the error was temporary, then after the final timeout, only
1266 "retry timeout exceeded" was returned. Now it returns the full error as
1267 well as "retry timeout exceeded".
1269 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1270 do this, and (what is worse) MTAs that accept it.
1272 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1273 will now be deprecated.
1275 PH/56 New os.c-cygwin from the Cygwin maintainer.
1277 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1278 formats (previously only available in exim formats via -bpr, -bpru,
1279 and -bpra. Now also available in native and exiqgrep formats)
1281 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1282 with very large, slow to parse queues
1284 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1286 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1288 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1289 responses to authentication challenges, though it was showing the
1290 challenges; (ii) I've removed the CR characters from the debug output for
1293 PH/58 Allow for the insertion of a newline as well as a space when a string
1294 is turned into more than one encoded-word during RFC 2047 encoding. The
1295 Sieve code now uses this.
1297 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1298 data_4xx, lost_connection, tls_required.
1300 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1301 message was being sent as an SMTP response.
1303 PH/61 Add -l and -k options to exicyclog.
1305 PH/62 When verifying, if an address was redirected to one new address, so that
1306 verification continued, and the new address failed or deferred after
1307 having set something in $address_data, the value of $address_data was not
1308 passed back to the ACL. This was different to the case when no
1309 redirection occurred. The value is now passed back in both cases.
1311 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1312 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1313 use login_cap.h, so on its own it isn't the distinguishing feature. The
1314 new name refers directly to the setclassresources() function.
1316 PH/65 Added configuration files for NetBSD3.
1318 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1320 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1321 is preferred over IPv4.
1323 PH/68 The bounce_return_message and bounce_return_body options were not being
1324 honoured for bounces generated during the reception of non-SMTP messages.
1325 In particular, this applied to messages rejected by the ACL. This bug has
1326 been fixed. However, if bounce_return_message is true and bounce_return_
1327 body is false, the headers that are returned for a non-SMTP message
1328 include only those that have been read before the error was detected.
1329 (In the case of an ACL rejection, they have all been read.)
1331 PH/69 The HTML version of the specification is now built in a directory called
1332 spec_html instead of spec.html, because the latter looks like a path with
1333 a MIME-type, and this confuses some software.
1335 PH/70 Catch two compiler warnings in sieve.c.
1337 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1338 function verify_get_ident() calls ip_connect() to connect a socket, but
1339 if the "connect()" function timed out, ip_connect() used to close the
1340 socket. However, verify_get_ident() also closes the socket later, and in
1341 between Exim writes to the log, which may get opened at this point. When
1342 the socket was closed in ip_connect(), the log could get the same file
1343 descriptor number as the socket. This naturally causes chaos. The fix is
1344 not to close the socket in ip_connect(); the socket should be closed by
1345 the function that creates it. There was only one place in the code where
1346 this was missing, in the iplookup router, which I don't think anybody now
1347 uses, but I've fixed it anyway.
1349 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1350 well as to direct DNS lookups. Otherwise the handling of names in host
1351 lists is inconsistent and therefore confusing.
1357 PH/01 Two changes to the default runtime configuration:
1359 (1) Move the checks for relay_from_hosts and authenticated clients from
1360 after to before the (commented out) DNS black list checks.
1362 (2) Add control=submission to the relay_from_hosts and authenticated
1363 clients checks, on the grounds that messages accepted by these
1364 statements are most likely to be submissions.
1366 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1368 (1) Generate an error if the third argument for the ${prvs expansion is
1371 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1374 (3) Reset the variables that are obtained from the first argument of
1375 ${prvscheck and used in the second argument before leaving the code,
1376 because their memory is reclaimed, so using them afterwards may do
1379 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1380 one (it's much easier than Tom thought :-).
1382 (5) Because of (4), we can now allow for the use of $prvscheck_result
1383 inside the third argument.
1385 PH/03 For some reason, the default setting of PATH when running a command from
1386 a pipe transport was just "/usr/bin". I have changed it to
1389 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1390 anything to be listed in the output from -bV.
1392 PH/05 When a filter generated an autoreply, the entire To: header line was
1393 quoted in the delivery log line, like this:
1395 => >A.N.Other <ano@some.domain> <original@ddress> ...
1397 This has been changed so that it extracts the operative address. There
1398 may be more than one such address. If so, they are comma-separated, like
1401 => >ano@some.domain,ona@other.domain <original@ddress> ...
1403 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1404 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1405 not being looked up in the rDNS to get a host name, Exim was showing the
1406 IP address twice in Received: lines, even though the IP addresses were
1407 identical. For example:
1409 Received: from [1.2.3.4] (helo=[1.2.3.4])
1411 However, if the real host name was known, it was omitting the HELO data
1412 if it matched the actual IP address. This has been tidied up so that it
1413 doesn't show the same IP address twice.
1415 PH/07 When both +timestamp and +memory debugging was on, the value given by
1416 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1417 called by the debug printing, thereby overwriting the timestamp buffer.
1418 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1420 PH/08 When the original message was included in an autoreply transport, it
1421 always said "this is a copy of the message, including all the headers",
1422 even if body_only or headers_only was set. It now gives an appropriate
1425 PH/09 Applied a patch from the Sieve maintainer which:
1427 o fixes some comments
1428 o adds the (disabled) notify extension core
1429 o adds some debug output for the result of if/elsif tests
1430 o points to the current vacation draft in the documentation
1431 and documents the missing references header update
1435 o fixes a bug in processing the envelope test (when testing
1436 multiple envelope elements, the last element determinted the
1439 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1440 Electronic Mail") by including:
1442 Auto-submitted: auto-generated
1444 in the messages that it generates (bounce messages and others, such as
1445 warnings). In the case of bounce messages for non-SMTP mesages, there was
1446 also a typo: it was using "Auto_submitted" (underscore instead of
1447 hyphen). Since every message generated by Exim is necessarily in response
1448 to another message, thes have all been changed to:
1450 Auto-Submitted: auto-replied
1452 in accordance with these statements in the RFC:
1454 The auto-replied keyword:
1456 - SHOULD be used on messages sent in direct response to another
1457 message by an automatic process,
1459 - MUST NOT be used on manually-generated messages,
1461 - MAY be used on Delivery Status Notifications (DSNs) and Message
1462 Disposition Notifications (MDNs),
1464 - MUST NOT be used on messages generated by automatic or periodic
1465 processes, except for messages which are automatic responses to
1468 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1469 to the default Received: header definition.
1471 PH/12 Added log selector acl_warn_skipped (default on).
1473 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1474 variables because (a) they are in the wrong storage pool and (b) even if
1475 they were copied, it wouldn't work properly because of the caching.
1477 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1478 checking when decoding. Apparently there are clients that generate
1479 overlong encoded strings. Why am I not surprised?
1481 PH/15 If the first argument of "${if match_address" was not empty, but did not
1482 contain an "@" character, Exim crashed. Now it writes a panic log message
1483 and treats the condition as false.
1485 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1487 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1488 extension "envelope-auth". The code is finished and in agreement with
1489 other implementations, but there is no documentation so far and in fact,
1490 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1491 not changing the active code.
1493 Print executed "if" and "elsif" statements when debugging is used. This
1494 helps a great deal to understand what a filter does.
1496 Document more things not specified clearly in RFC3028. I had all this
1497 sorted out, when out of a sudden new issues came to my mind. Oops."
1499 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1502 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1503 canonical form (as documented). However, after a host name lookup from
1504 the IP address, check_host() was doing a simple string comparison with
1505 addresses acquired from the DNS when checking that the found name did
1506 have the original IP as one of its addresses. Since any found IPv6
1507 addresses are likely to be in abbreviated form, the comparison could
1508 fail. Luckily, there already exists a function for doing the comparison
1509 by converting both addresses to binary, so now that is used instead of
1510 the text comparison.
1512 PH/20 There was another similar case to PH/19, when a complete host name was
1513 given in a host list; looking up its IP address could give an abbreviated
1514 form, whereas the current host's name might or might not be abbreviated.
1515 The same fix has been applied.
1521 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1522 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1525 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1526 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1528 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1530 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1531 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1532 Also, for "Auto-Submitted", treat anything other than "no" as
1533 non-personal, in accordance with RFC 3834. (Previously it treated
1534 anything starting "auto-" as non-personal.)
1536 TF/01 The control=submission/name=... option had a problem with syntax
1537 errors if the name included a slash character. The /name= option
1538 now slurps the rest of the string, so it can include any characters
1539 but it must come last in the list of options (after /sender_retain
1542 PH/05 Some modifications to the interface to the fake nameserver for the new
1550 TK/01 Added the "success_on_redirect" address verification option. See
1551 NewStuff for rationale and an example.
1553 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1555 PH/02 Patch to exigrep to allow it to work on syslog lines.
1557 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1558 fread() to skip over the body file's header line, because in Cygwin the
1559 header line is locked and is inaccessible.
1561 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1562 co-exist for some time) to make it clear that it is the Exim ID that is
1563 referenced, not the Message-ID: header line.
1565 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1566 string_format() function, because snprintf() does not exist on all
1569 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1570 ${stat: expansion item.
1572 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1573 protocol synchronization error", to keep the pedants happy.
1575 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1576 well as for IRIX systems, when gcc is being used. See the host.c source
1579 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1581 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1584 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1585 transport and to the smtp transport in LMTP mode.
1587 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1589 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1590 run a filter in a subprocess. This could lead to confusion in subsequent
1591 lookups in the parent process. There should also be a search_tidyup() at
1592 the end of the subprocess.
1594 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1595 only if the host matched helo_try_verify_hosts, which caused the
1596 verification to occur when the EHLO/HELO command was issued. The ACL just
1597 tested the remembered result. Now, if a previous verification attempt has
1598 not happened, "verify = helo" does it there and then.
1600 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1602 TK/03 Fix log output including CR from clamd.
1604 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1605 exist provoked a memory error which could cause a segfault.
1607 PH/15 Installed PCRE 6.2
1609 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1611 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1612 of the problem. Specifically, suggested +O2 rather than +O1 for the
1615 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1617 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1618 record, the port was not correctly logged when the outgoing_port log
1619 selector was set (it logged the transort's default port).
1621 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1622 fallback_hosts, and "hosts" in the smtp transport.
1624 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1625 host errors such as "Connection refused".
1627 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1628 authentication with radiusclient 0.4.9:
1630 - Error returned from rc_read_config was caught wrongly
1631 - Username/password not passed on to radius server due to wrong length.
1633 The presumption is that some radiusclient API changes for 4.51/PH/17
1634 were not taken care of correctly. The code is still untested by me (my
1635 Linux distribution still has 0.3.2 of radiusclient), but it was
1636 contributed by a Radius user.
1638 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1639 expanding the "port" option of the smtp transport.
1641 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1642 while reading a MIME header. Thanks to Tom Hughes for a patch.
1644 PH/24 Include config.h inside local_scan.h so that configuration settings are
1647 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1648 that in an ACL for RCPT (for example), you can examine exactly what was
1651 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1652 commands, but it was not correctly comparing the address with the actual
1653 client host address. Thus, it would show the EHLO address in Received:
1654 header lines when this was not necessary.
1656 PH/27 Added the % operator to ${eval:}.
1658 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1659 it should be ignoring failures (because with -C, for example, it has lost
1660 privilege). It wasn't ignoring creation failures other than "already
1663 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1666 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1667 because an input line was too long, either on its own, or by virtue of
1668 too many continuations, the temporary file was not being removed, and the
1669 return code was incorrect.
1671 PH/31 Missing "BOOL" in function definition in filtertest.c.
1673 PH/32 Applied Sieve patches from the maintainer.
1675 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1677 PH/33 Added "verify = not_blind".
1679 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1680 Local/Makefile (with some defaults set). These are used in built scripts
1681 such as exicyclog, but they have never been used in the exim_install
1682 script (though there are many overriding facilities there). I have
1683 arranged that the exim_install script now takes note of these two
1686 PH/35 Installed configuration files for Dragonfly.
1688 PH/36 When a locally submitted message by a trusted user did not contain a
1689 From: header, and the sender address was obtained from -f or from an SMTP
1690 MAIL command, and the trusted user did not use -F to supply a sender
1691 name, $originator_name was incorrectly used when constructing a From:
1692 header. Furthermore, $originator_name was used for submission mode
1693 messages from external hosts without From: headers in a similar way,
1694 which is clearly wrong.
1696 PH/37 Added control=suppress_local_fixups.
1698 PH/38 When log_selector = +received_sender was set, and the addition of the
1699 sender made the log line's construction buffer exactly full, or one byte
1700 less than full, an overflow happened when the terminating "\n" was
1703 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1704 when the result of a list match is failure because a DNS lookup failed.
1706 PH/40 RM_COMMAND is now used in the building process.
1708 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1709 the "build-* directories that it finds.
1711 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1712 address in a domain literal was a prefix of an interface address.
1714 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1715 when verifying a sender address, unless rewrite_headers is false.
1717 PH/44 Wrote a long comment about why errors_to addresses are verified as
1718 recipients, not senders.
1720 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1721 the ratelimit ACL was added.
1723 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1725 PH/46 Added extra information about PostgreSQL errors to the error string.
1727 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1728 avoiding the need to install special zones in a real server. This is
1729 backwards compatible; if it can't find the fake resolver, it drops back.
1730 Thus, both old and new test suites can be run.
1732 TF/02 Added util/ratelimit.pl
1734 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1735 clock is set back in time.
1737 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1738 Candler <B.Candler@pobox.com>.
1740 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1741 OK for addresses that are the result of redirections.
1743 PH/48 A number of further additions for the benefit of the new test suite,
1744 including a fake gethostbyname() that interfaces to the fake DNS resolver
1747 TF/06 The fix for widen_domains has also been applied to qualify_single and
1748 search_parents which are the other dnslookup options that can cause
1751 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1754 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1756 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1757 service" if it encountered an MX record with an empty target hostname.
1758 The message is now "an MX or SRV record indicated no SMTP service".
1760 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1761 if the DNS of the sending site is misconfigured. This is quite a
1762 common situation. This change restores the behaviour of treating a
1763 helo verification defer as a failure.
1765 PH/51 If self=fail was set on a router, the bounce message did not include the
1766 actual error message.
1772 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1774 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1775 command itself ended in error, the underlying message about the transport
1776 filter timeout was being overwritten with the pipe command error. Now the
1777 underlying error message should be appended to the second error message.
1779 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1781 PH/02 Reduce the amount of output that "make" produces by default. Full output
1782 can still be requested.
1784 PH/03 The warning log line about a condition test deferring for a "warn" verb
1785 was being output only once per connection, rather than after each
1786 occurrence (because it was using the same function as for successful
1787 "warn" verbs). This seems wrong, so I have changed it.
1789 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1790 it should not have, which might have caused a crash in the right
1791 circumstances, but probably never did.
1793 PH/04 Installed a modified version of Tony Finch's patch to make submission
1794 mode fix the return path as well as the Sender: header line, and to
1795 add a /name= option so that you can make the user's friendly name appear
1798 TF/03 Added the control = fakedefer ACL modifier.
1800 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1801 Mark Lowes for thorough testing.
1803 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1805 TK/03 Merged latest SRS patch from Miles Wilton.
1807 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1808 with the definition in sysexits.h (which is #included earlier).
1809 Fortunately, Exim does not actually use EX_OK. The code used to try to
1810 preserve the sysexits.h value, by assumimg that macro definitions were
1811 scanned for macro replacements. I have been disabused of this notion,
1812 so now the code just undefines EX_OK before #including unistd.h.
1814 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1815 in the smtp transport. When a block could not be written in a single
1816 write() function, the timeout was being re-applied to each part-write.
1817 This seems wrong - if the receiver was accepting one byte at a time it
1818 would take for ever. The timeout is now adjusted when this happens. It
1819 doesn't have to be particularly precise.
1821 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1822 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1824 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1825 without a domain if the check to <postmaster@domain> fails.
1827 SC/01 Eximstats: added -xls and the ability to specify output files
1828 (patch written by Frank Heydlauf).
1830 SC/02 Eximstats: use FileHandles for outputing results.
1832 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1834 SC/04 Eximstats: fixed display of large numbers with -nvr option
1836 SC/05 Eximstats: fixed merging of reports with empty tables.
1838 SC/06 Eximstats: added the -include_original_destination flag
1840 SC/07 Eximstats: removed tabs and trailing whitespace.
1842 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1844 TK/06 MBOX spool code: Add real "From " MBOX separator line
1845 so the .eml file is really in mbox format (even though
1846 most programs do not really care). Patch from Alex Miller.
1848 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1849 The latter is generated from $received_to and is only set if the
1850 message has one envelope recipient. SA can use these headers,
1851 obviously out-of-the-box. Patch from Alex Miller.
1853 PH/08 The ${def test on a variable was returning false if the variable's
1854 value was "0", contrary to what the specification has always said!
1855 The result should be true unless the variable is empty.
1857 PH/09 The syntax error of a character other than { following "${if
1858 def:variable_name" (after optional whitespace) was not being diagnosed.
1859 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1860 accidental colon was present, for example, could give incorrect results.
1862 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1863 result is used (not including appendfile, where other changes are about
1866 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1867 This involved changing a lot of size variables from int to off_t. It
1868 should work with maildirs and everything.
1870 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1871 spamd dying while we are connected to it.
1873 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1876 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1877 function for BDB 4.3.
1879 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1881 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1882 the sources for PCRE have been re-arranged and more files are now
1885 PH/15 The code I had for printing potentially long long variables in PH/11
1886 above was not the best (it lost precision). The length of off_t variables
1887 is now inspected at build time, and an appropriate printing format (%ld
1888 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1889 to be "long long int" or "long int". This is needed for the internal
1890 formatting function string_vformat().
1892 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1893 the configuration file to be ":syslog", then the script "guesses" where
1894 the logs files are, rather than using the compiled in default. In our
1895 case the guess is not the same as the compiled default, so the script
1896 suddenly stopped working when I started to use syslog. The patch checks
1897 to see if log_file_path is "". If so, it attempts to read it from exim
1898 with no configuration file to get the compiled in version, before it
1899 falls back to the previous guessing code."
1901 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1902 implementing BATV in an Exim configuration. See NewStuff for the gory
1905 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1906 Makefile that are specific to HP-UX.
1908 PH/18 If the "use_postmaster" option was set for a recipient callout together
1909 with the "random" option, the postmaster address was used as the MAIL
1910 FROM address for the random test, but not for the subsequent recipient
1911 test. It is now used for both.
1913 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1914 patch removes a few documentation additions to RFC 3028, because the
1915 latest draft now contains them. It adds the new en;ascii-case comparator
1916 and a new error check for 8bit text in MIME parts. Comparator and
1917 require names are now matched exactly. I enabled the subaddress
1918 extension, but it is not well tested yet (read: it works for me)."
1920 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1921 rework some of the code of TK/09 above to avoid the hardwired use of
1922 "%lld" and "long long". Replaced the call to snprintf() with a call to
1925 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1926 records point to non-existent hosts", "retry timeout exceeded", and
1927 "retry time not reached for any host after a long failure period".
1929 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1930 experimental DomainKeys support:
1932 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1933 (2) On an error such as an illegally used "control", the wrong name for
1934 the control was given.
1936 These problems did NOT occur unless DomainKeys support was compiled.
1938 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1940 PH/24 Added ${if match_ip condition.
1942 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1943 databases so that it will be absolutely obvious if a crash occurs in the
1944 DB library. This is a regular occurrence (often caused by mis-matched
1947 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1948 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1949 picked up on a user's system that detects such things. There doesn't seem
1950 to be a gcc warning option for this - only an attribute that has to be
1951 put on the function's prototype. It seems that in Fedora Core 4 they have
1952 set this on a number of new functions. No doubt there will be more in due
1955 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1956 specify a transport. However, if an address that was verified by such a
1957 router was the subject of a callout, Exim crashed because it tried to
1958 read the rcpt_include_affixes from the non-existent transport. Now it
1959 just assumes that the setting of that option is false. This bug was
1960 introduced by 4.51/PH/31.
1962 PH/28 Changed -d+all to exclude +memory, because that information is very
1963 rarely of interest, but it makes the output a lot bigger. People tend to
1964 do -d+all out of habit.
1966 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1967 code in os-type was giving problems when libc.so lives in lib64, like on
1970 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1971 aren't the modern standard, and it seems that some systems' include files
1972 don't always have them. Exim was already checking for some of the newer
1973 ones like T_AAAA, and defining it itself. I've added checks for all the
1974 record types that Exim uses.
1976 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1977 not automatically generating a new one, as it is supposed to. This
1978 prevented TLS from working. If the file did exist, but contained invalid
1979 data, a new version was generated, as expected. It was only the case of a
1980 non-existent file that was broken.
1982 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1983 with a change in libdomainkeys > 0.64.
1985 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1986 from DNS. If the selector record carries the flag, it now has
1987 precedence over the domain-wide flag.
1989 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1991 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1992 the use of a port name that isn't defined in /etc/services) occurred, the
1993 message was deferred as in a normal delivery, and thus remained on the
1994 spool, instead of being failed because of the mua_wrapper setting. This
1995 is now fixed, and I tidied up some of the mua_wrapper messages at the
1998 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1999 the messages in a hash of arrays rather than using individual hashes.
2000 This is a bit cleaner and results in dramatic memory savings, albeit
2001 at a slight CPU cost.
2003 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2004 as requested by Marc Sherman.
2006 SC/10 Eximstats: added histograms for user specified patterns as requested
2009 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2011 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2012 fopen() in the content-scanning modules that did not already have it.
2018 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2019 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2021 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2023 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2025 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2026 to test Sieve filters that use "vacation".
2028 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2029 that changes the way the GnuTLS parameters are stored in the cache file.
2030 The new format can be generated externally. For backward compatibility,
2031 if the data in the cache doesn't make sense, Exim assumes it has read an
2032 old-format file, and it generates new data and writes a new file. This
2033 means that you can't go back to an older release without removing the
2036 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2037 work if there are any delivery delays because "one_time" forces the
2038 parent to be marked "delivered", so its unseen clone is never tried
2039 again. For this reason, Exim now forbids the simultaneous setting of
2042 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2043 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2044 entry if you want to know the details. The fix, however, neglected to
2045 consider the case when local delivery batching is involved. The test for
2046 "previously delivered" was not happening when checking to see if an
2047 address could be batched with a previous (undelivered) one; under
2048 certain circumstances this could lead to multiple deliveries to the same
2051 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2052 in its include files, and this causes problems building Exim.
2054 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2055 header_syntax) but Exim was just ignoring anything given after a slash.
2056 In particular, this caused confusion with an attempt to use "verify =
2057 reverse_host_lookup/defer_ok". An error is now given when options are
2058 supplied for verify items that do not have them. (Maybe reverse_host_
2059 lookup should have a defer_ok option, but that's a different point.)
2061 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2062 defined by RFC 821) to 2048, because there were problems with some AUTH
2063 commands, and RFC 1869 says the size should be increased for extended
2064 SMTP commands that take arguments.
2066 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2069 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2070 "unknown" error; now it says that the functionality isn't in the binary.
2072 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2073 an address' error message when a string expansion fails (syntax or
2074 whatever). Otherwise the password may appear in the log. Following change
2075 PH/42 below, there is no longer a chance of it appearing in a bounce
2078 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2080 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2081 of its addresses was ever tried. (Bugzilla bug #2.)
2083 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2084 the result incorrectly in the debug output. (It correctly added a newline
2085 to what was transported.)
2087 TF/01 Added $received_time.
2089 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2090 commented out examples of how to interface to a virus scanner and to
2091 SpamAssassin. Also added commented examples of av_scanner and
2092 spamd_address settings.
2094 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2095 and controls are allowed in which ACLs. There were a couple of minor
2096 errors. Some of the entries in the conditions table (which is a table of
2097 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2098 negation of where the condition IS allowed.
2100 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2102 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2103 header file does not have a version number, so I've had to invent a new
2104 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2105 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2106 radiusclient), but it was contributed by a Radius user.
2108 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2109 files or directories, for OpenSSL.
2111 PH/19 When an Exim process that is running as root has to create an Exim log
2112 file, it does so in a subprocess that runs as exim:exim so as to get the
2113 ownership right at creation (otherwise, other Exim processes might see
2114 the file with the wrong ownership). There was no test for failure of this
2115 fork() call, which would lead to the process getting stuck as it waited
2116 for a non-existent subprocess. Forks do occasionally fail when resources
2117 run out. I reviewed all the other calls to fork(); they all seem to check
2120 PH/20 When checking for unexpected SMTP input at connect time (before writing
2121 the banner), Exim was not dealing correctly with a non-positive return
2122 from the read() function. If the client had disconnected by this time,
2123 the result was a log entry for a synchronization error with an empty
2124 string after "input=" when read() returned zero. If read() returned -1
2125 (an event I could not check), uninitialized data bytes were printed.
2126 There were reports of junk text (parts of files, etc) appearing after
2129 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2131 PH/22 Added support for macro redefinition, and (re)definition in between
2132 driver and ACL definitions.
2134 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2135 forgetting to use the resulting value; it was using the unexpanded value.
2137 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2138 hadn't been configured. The fix is from Juergen Kreileder, who
2139 understands it better than I do:
2141 "Here's what I see happening with three configured cyrus_sasl
2142 authenticators configured (plain, login, cram-md5):
2144 On startup auth_cyrus_sasl_init() gets called for each of these.
2145 This means three calls to sasl_listmech() without a specified mech_list.
2146 => SASL tests which mechs of all available mechs actually work
2147 => three warnings about OTP not working
2148 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2150 With the patch, sasl_listmech() also gets called three times. But now
2151 SASL's mech_list option is set to the server_mech specified in the the
2152 authenticator. Or in other words, the answer from sasl_listmech()
2153 gets limited to just the mech you're testing for (which is different
2155 => the return list contains just 'plain' or 'login', 'cram-md5' or
2156 nothing depending on the value of ob->server_mech.
2158 I've just tested the patch: Authentication still works fine,
2159 unavailable mechs specified in the exim configuration are still
2160 caught, and the auth.log warnings about OTP are gone."
2162 PH/25 When debugging is enabled, the contents of the command line are added
2163 to the debugging output, even when log_selector=+arguments is not
2166 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2167 answer is "GNU", and only if the return is "GNU/something" is the answer
2170 PH/27 $acl_verify_message is now set immediately after the failure of a
2171 verification in an ACL, and so is available in subsequent modifiers. In
2172 particular, the message can be preserved by coding like this:
2174 warn !verify = sender
2175 set acl_m0 = $acl_verify_message
2177 Previously, $acl_verify_message was set only while expanding "message"
2178 and "log_message" when a very denied access.
2180 PH/28 Modified OS/os.c-Linux with
2182 -#ifndef OS_LOAD_AVERAGE
2183 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2185 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2186 nomenclature these days.)
2188 PH/29 Installed patch from the Sieve maintainer that adds the options
2189 sieve_useraddress and sieve_subaddress to the redirect router.
2191 PH/30 In these circumstances:
2192 . Two addresses routed to the same list of hosts;
2193 . First host does not offer TLS;
2194 . First host accepts first address;
2195 . First host gives temporary error to second address;
2196 . Second host offers TLS and a TLS session is established;
2197 . Second host accepts second address.
2198 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2199 and peerdn, if requested) that were in fact used only for the second
2202 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2203 attention to any local part prefix or suffix that was matched by the
2204 router that accepted the address. It now behaves in the same way as it
2205 does for delivery: the affixes are removed from the local part unless
2206 rcpt_include_affixes is set on the transport.
2208 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2209 timeout during the DATA phase of an incoming message.
2211 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2212 applied a patch sanctioned by the Sieve maintainer.
2214 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2215 the uid or gid is negative. A case of a negative gid caused this to be
2216 noticed. The fix allows for either to be negative.
2218 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2219 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2222 PH/36 The change PH/12 above was broken. Fixed it.
2224 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2225 the grounds that routing the same address twice would always produce the
2226 same answer. This might have been true once, but it is certainly no
2227 longer true now. Routing a child address may depend on the previous
2228 routing that produced that child. Some complicated redirection strategies
2229 went wrong when messages had multiple recipients, and made Exim's
2230 behaviour dependent on the order in which the addresses were given.
2232 I have moved the duplicate checking until after the routing is complete.
2233 Exim scans the addresses that are assigned to local and remote
2234 transports, and removes any duplicates. This means that more work will be
2235 done, as duplicates will always all be routed, but duplicates are
2236 presumably rare, so I don't expect this is of any significance.
2238 For deliveries to pipes, files, and autoreplies, the duplicate checking
2239 still happens during the routing process, since they are not going to be
2242 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2243 It corrects a timeout issue with spamd. This is Ian's comment: "The
2244 background is that sometimes spamd either never reads data from a
2245 connection it has accepted, or it never writes response data. The exiscan
2246 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2247 blindly assumes that writes won't block so it may never time out."
2249 PH/39 Allow G after quota size as well as K and M.
2251 PH/40 The value set for $authenticated_id in an authenticator may not contain
2252 binary zeroes or newlines because the value is written to log lines and
2253 to spool files. There was no check on this. Now the value is run through
2254 the string_printing() function so that such characters are converted to
2255 printable escape sequences.
2257 PH/41 $message_linecount is a new variable that contains the total number of
2258 lines in the message. Compare $body_linecount, which is the count for the
2261 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2262 bounce and delay warning messages, except in certain special cases, which
2265 (a) An SMTP error message from a remote host;
2266 (b) A message specified in a :fail: redirection;
2267 (c) A message specified in a "fail" command in a system filter;
2268 (d) A message specified in a FAIL return from the queryprogram router;
2269 (e) A message specified by the cannot_route_message router option.
2271 In these cases only, Exim does include the error details in bounce and
2272 warning messages. There are also a few cases where bland messages such
2273 as "unrouteable address" or "local delivery error" are given.
2275 PH/43 $value is now also set for the "else" part of a ${run expansion.
2277 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2278 being worked on, but at least Exim now implements the latest version to
2281 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2282 process to complete was treated as a delivery failure, a timeout while
2283 writing the message to the pipe was logged, but erroneously treated as a
2284 successful delivery. Such timeouts include transport filter timeouts. For
2285 consistency with the overall process timeout, these timeouts are now
2286 treated as errors, giving rise to delivery failures by default. However,
2287 there is now a new Boolean option for the pipe transport called
2288 timeout_defer, which, if set TRUE, converts the failures into defers for
2289 both kinds of timeout. A transport filter timeout is now identified in
2292 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2293 systems where "make" and "gmake" are different, calling "gmake" at top
2294 level broke things. I've arranged for the value of $(MAKE) to be passed
2295 from the Makefile to this script so that it can call the same version of
2299 A note about Exim versions 4.44 and 4.50
2300 ----------------------------------------
2302 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2303 changes of various kinds. As a consequence, a big documentation update was
2304 needed. This delayed the release for rather longer than seemed good, especially
2305 in the light of a couple of (minor) security issues. Therefore, the changes
2306 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2307 release. So 4.44 and 4.50 are in effect two different branches that both start
2310 I have left the 4.50 change log unchanged; it contains all the changes since
2311 4.43. The change log for 4.44 is below; many of its items are identical to
2312 those for 4.50. This seems to be the most sensible way to preserve the
2313 historical information.
2319 1. Minor wording change to the doc/README.SIEVE file.
2321 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2322 computation of the current number of files was incorrect.
2324 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2325 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2328 4. Give more explanation in the error message when the command for a transport
2329 filter fails to execute.
2331 5. There are several places where Exim runs a non-Exim command in a
2332 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2333 was being done only for the command run by the queryprogram router. It is
2334 now done for all such subprocesses. The other cases are: ${run, transport
2335 filters, and the commands run by the lmtp and pipe transports.
2337 6. Added CONFIGURE_GROUP build-time option.
2339 7. Some older OS have a limit of 256 on the maximum number of file
2340 descriptors. Exim was using setrlimit() to set 1000 as a large value
2341 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2342 systems. I've change it so that if it can't get 1000, it tries for 256.
2344 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2345 was an oversight, and furthermore, ever since the addition of extra
2346 controls (e.g. 4.43/32), the checks on when to allow different forms of
2347 "control" were broken. There should now be diagnostics for all cases when a
2348 control that does not make sense is encountered.
2350 9. Added the /retain_sender option to "control=submission".
2352 10. $recipients is now available in the predata ACL (oversight).
2354 11. Tidy the search cache before the fork to do a delivery from a message
2355 received from the command line. Otherwise the child will trigger a lookup
2356 failure and thereby defer the delivery if it tries to use (for example) a
2357 cached ldap connection that the parent has called unbind on.
2359 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2360 of $address_data from the recipient verification was clobbered by the
2361 sender verification.
2363 13. The value of address_data from a sender verification is now available in
2364 $sender_address_data in subsequent conditions in the ACL statement.
2366 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2368 15. Added a new option "connect=<time>" to callout options, to set a different
2371 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2372 was its contents. (It was OK if the option was not defined at all.)
2374 17. A "Completed" log line is now written for messages that are removed from
2375 the spool by the -Mrm option.
2377 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2378 information about exactly what failed.
2380 19. Added -dd to debug only the daemon process.
2382 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2383 handles timeouts, both on the server side and network timeouts. Renamed the
2384 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2386 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2387 It is now set to "smtps".
2389 22. $host_address is now set to the target address during the checking of
2390 ignore_target_hosts.
2392 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2393 being passed; this would have caused $sender_host_name to have been used if
2394 matching the list had actually called for a host name (not very likely,
2395 since this list is usually IP addresses). A host name is now passed as
2398 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2399 code that specifies a non-restarting handler (typically sigaction() in
2400 modern systems) in an attempt to fix a rare and obscure crash bug.
2402 25. Narrowed the window for a race in the daemon that could cause it to ignore
2403 SIGCHLD signals. This is not a major problem, because they are used only to
2404 wake it up if nothing else does.
2406 26. A malformed maildirsize file could cause Exim to calculate negative values
2407 for the mailbox size or file count. Odd effects could occur as a result.
2408 The maildirsize information is now recalculated if the size or filecount
2411 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2412 support for a long time. Removed HAVE_SYS_VFS_H.
2414 28. Installed the latest version of exipick from John Jetmore.
2416 29. In an address list, if the pattern was not a regular expression, an empty
2417 subject address (from a bounce message) matched only if the pattern was an
2418 empty string. Non-empty patterns were not even tested. This was the wrong
2419 because it is perfectly reasonable to use an empty address as part of a
2420 database query. An empty address is now tested by patterns that are
2421 lookups. However, all the other forms of pattern expect the subject to
2422 contain a local part and a domain, and therefore, for them, an empty
2423 address still always fails if the pattern is not itself empty.
2425 30. Exim went into a mad DNS loop when attempting to do a callout where the
2426 host was specified on an smtp transport, and looking it up yielded more
2427 than one IP address.
2429 31. Re-factored the code for checking spool and log partition space into a
2430 function that finds that data and another that does the check. The former
2431 is then used to implement four new variables: $spool_space, $log_space,
2432 $spool_inodes, and $log_inodes.
2434 32. The RFC2047 encoding function was originally intended for short strings
2435 such as real names; it was not keeping to the 75-character limit for
2436 encoded words that the RFC imposes. It now respects the limit, and
2437 generates multiple encoded words if necessary. To be on the safe side, I
2438 have increased the buffer size for the ${rfc2047: expansion operator from
2441 33. It is now permitted to omit both strings after an "if" condition; if the
2442 condition is true, the result is "true". As before, when the second string
2443 is omitted, a false condition yields an empty string. This makes it less
2444 cumbersome to write custom ACL and router conditions.
2446 34. Failure to deliver a bounce message always caused it to be frozen, even if
2447 there was an errors_to setting on the router. The errors_to setting is now
2450 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2451 canonical form (fully expanded) before being placed in
2452 $sender_host_address.
2454 36. The table in the code that translates DNS record types into text (T_A to
2455 "A" for instance) was missing entries for NS and CNAME. It is just possible
2456 that this could have caused confusion if both these types were looked up
2457 for the same domain, because the text type is used as part of Exim's
2458 per-process caching. But the chance of anyone hitting this buglet seems
2461 37. The dnsdb lookup has been extended in a number of ways.
2463 (1) There is a new type, "zns", which walks up the domain tree until it
2464 finds some nameserver records. It should be used with care.
2466 (2) There is a new type, "mxh", which is like "mx" except that it returns
2467 just the host names, not the priorities.
2469 (3) It is now possible to give a list of domains (or IP addresses) to be
2470 looked up. The behaviour when one of the lookups defers can be
2471 controlled by a keyword.
2473 (4) It is now possible to specify the separator character for use when
2474 multiple records are returned.
2476 38. The dnslists ACL condition has been extended: it is now possible to supply
2477 a list of IP addresses and/or domains to be looked up in a particular DNS
2480 39. Added log_selector=+queue_time_overall.
2482 40. When running the queue in the test harness, wait just a tad after forking a
2483 delivery process, to get repeatability of debugging output.
2485 41. Include certificate and key file names in error message when GnuTLS fails
2486 to set them up, because the GnuTLS error message doesn't include the name
2487 of the failing file when there is a problem reading it.
2489 42. Allow both -bf and -bF in the same test run.
2491 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2493 44. The "Exiscan patch" is now merged into the mainline Exim source.
2495 45. Sometimes the final signoff response after QUIT could fail to get
2496 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2497 < 0 before doing a fflush(). This bug looks as though it goes back to the
2498 introduction of TLS in release 3.20, but "sometimes" must have been rare
2499 because the tests only now provoked it.
2501 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2502 (this can affect the format of dates).
2504 47. exim_tidydb, when checking for the continued existence of a message for
2505 which it has found a message-specific retry record, was not finding
2506 messages that were in split spool directories. Consequently, it was
2507 deleting retry records that should have stayed in existence.
2509 48. Steve fixed some bugs in eximstats.
2511 49. The SPA authentication driver was not abandoning authentication and moving
2512 on to the next authenticator when an expansion was forced to fail,
2513 contradicting the general specification for all authenticators. Instead it
2514 was generating a temporary error. It now behaves as specified.
2516 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2517 (the order specifies the preference for clients). The order is now AES256,
2518 AES128, 3DES, ARCFOUR128.
2520 51. Small patch to Sieve code - explicitly set From: when generating an
2523 52. Exim crashed if a remote delivery caused a very long error message to be
2524 recorded - for instance if somebody sent an entire SpamAssassin report back
2525 as a large number of 550 error lines. This bug was coincidentally fixed by
2526 increasing the size of one of Exim's internal buffers (big_buffer) that
2527 happened as part of the Exiscan merge. However, to be on the safe side, I
2528 have made the code more robust (and fixed the comments that describe what
2531 53. Now that there can be additional text after "Completed" in log lines (if
2532 the queue_time_overall log selector is set), a one-byte patch to exigrep
2533 was needed to allow it to recognize "Completed" as not the last thing in
2536 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2537 patch that reportedly fixes this has been added. I am not expert enough to
2538 create a test for it. This is what the patch creator wrote:
2540 "I found a little strange behaviour of ldap code when working with
2541 Windows 2003 AD Domain, where users was placed in more than one
2542 Organization Units. When I tried to give exim partial DN, the exit code
2543 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2544 But simultaneously result of request was absolutely normal ldap result,
2545 so I produce this patch..."
2547 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2548 REFERENCE, so I have modified the code to exclude the patch when that macro
2551 55. Some experimental protocols are using DNS PTR records for new purposes. The
2552 keys for these records are domain names, not reversed IP addresses. The
2553 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2554 leaves it alone. Component reversal etc. now happens only for IP addresses.
2557 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2559 57. Double the size of the debug message buffer (to 2048) so that more of very
2560 long debug lines gets shown.
2562 58. The exicyclog utility now does better if the number of log files to keep
2563 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2565 59. Two changes related to the smtp_active_hostname option:
2567 (1) $smtp_active_hostname is now available as a variable.
2568 (2) The default for smtp_banner uses $smtp_active_hostname instead
2569 of $primary_hostname.
2571 60. The host_aton() function is supposed to be passed a string that is known
2572 to be a valid IP address. However, in the case of IPv6 addresses, it was
2573 not checking this. This is a hostage to fortune. Exim now panics and dies
2574 if the condition is not met. A case was found where this could be provoked
2575 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2576 components; fortuitously, this particular loophole had already been fixed
2577 by change 4.50/55 above.
2579 If there are any other similar loopholes, the new check in host_aton()
2580 itself should stop them being exploited. The report I received stated that
2581 data on the command line could provoke the exploit when Exim was running as
2582 exim, but did not say which command line option was involved. All I could
2583 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2584 running as the user.
2587 61. There was a buffer overflow vulnerability in the SPA authentication code
2588 (which came originally from the Samba project). I have added a test to the
2589 spa_base64_to_bits() function which I hope fixes it.
2592 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2593 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2595 63. The daemon start-up calls getloadavg() while still root for those OS that
2596 need the first call to be done as root, but it missed one case: when
2597 deliver_queue_load_max is set with deliver_drop_privilege. This is
2598 necessary for the benefit of the queue runner, because there is no re-exec
2599 when deliver_drop_privilege is set.
2601 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2602 This has been fixed.
2604 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2605 was in use, was not putting the data itself into the right store pool;
2606 consequently, it could be overwritten for a subsequent message in the same
2607 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2610 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2612 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2613 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2614 the function were treating the return as a boolean value, which happened to
2615 work because 0=false and not-0=true, but is not correct code.
2617 68. The host_aton() function was not handling scoped IPv6 addresses (those
2618 with, for example, "%eth0" on the end) correctly.
2620 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2621 negated items (that is, ~something) in unsigned ints. Some compilers
2622 apparently mutter when there is no cast.
2624 70. If an address verification called from an ACL failed, and did not produce a
2625 user-specific message (i.e. there was only a "system" message), nothing was
2626 put in $acl_verify_message. In this situation, it now puts the system
2629 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2630 session; change 4.31/43 added the unwanted input to the log line - except
2631 that it did not do this in the start of session case. It now does.
2633 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2634 This is wrong and can cause the other end to generate a synchronization
2635 error if it is another Exim or anything else that does the synchronization
2636 check. A QUIT command is no longer sent after a timeout.
2638 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2639 during host lookups.
2641 74. The defer_ok option of callout verification was not working if it was used
2642 when verifying addresses in header lines, that is, for this case:
2644 verify = header_sender/callout=defer_ok
2646 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2647 those file descriptors could be used for SMTP connections. If anything
2648 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2649 could be sent to the SMTP client, causing chaos. The daemon now opens
2650 stdin, stdout, and stderr to /dev/null when it puts itself into the
2653 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2654 log by default. The user can override this with suitable Perl magic.
2656 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2657 the log message when discard triggers, was not working for the DATA ACL or
2658 for the non-SMTP ACL.
2660 78. Error message wording change in sieve.c.
2662 79. If smtp_accept_max_per_host was set, the number of connections could be
2663 restricted to fewer than expected, because the daemon was trying to set up
2664 a new connection before checking whether the processes handling previous
2665 connections had finished. The check for completed processes is now done
2666 earlier. On busy systems, this bug wouldn't be noticed because something
2667 else would have woken the daemon, and it would have reaped the completed
2670 80. If a message was submitted locally by a user whose login name contained one
2671 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2672 It caused a spool format error. I have fixed the spool reading code. A
2673 related problem was that the "from" clause in the Received: line became
2674 illegal because of the space(s). It is now covered by ${quote_local_part.
2676 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2679 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2681 83. Patch from Sieve maintainer to fix unterminated string problem in
2682 "vacation" handling.
2684 84. Some minor changes to the Linux configuration files to help with other
2685 OS variants using glibc.
2687 85. One more patch for Sieve to update vacation handling to latest spec.
2690 ----------------------------------------------------
2691 See the note above about the 4.44 and 4.50 releases.
2692 ----------------------------------------------------
2698 1. Change 4.43/35 introduced a bug that caused file counts to be
2699 incorrectly computed when quota_filecount was set in an appendfile
2702 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2703 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2706 3. Give more explanation in the error message when the command for a transport
2707 filter fails to execute.
2709 4. There are several places where Exim runs a non-Exim command in a
2710 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2711 was being done only for the command run by the queryprogram router. It is
2712 now done for all such subprocesses. The other cases are: ${run, transport
2713 filters, and the commands run by the lmtp and pipe transports.
2715 5. Some older OS have a limit of 256 on the maximum number of file
2716 descriptors. Exim was using setrlimit() to set 1000 as a large value
2717 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2718 systems. I've change it so that if it can't get 1000, it tries for 256.
2720 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2721 was an oversight, and furthermore, ever since the addition of extra
2722 controls (e.g. 4.43/32), the checks on when to allow different forms of
2723 "control" were broken. There should now be diagnostics for all cases when a
2724 control that does not make sense is encountered.
2726 7. $recipients is now available in the predata ACL (oversight).
2728 8. Tidy the search cache before the fork to do a delivery from a message
2729 received from the command line. Otherwise the child will trigger a lookup
2730 failure and thereby defer the delivery if it tries to use (for example) a
2731 cached ldap connection that the parent has called unbind on.
2733 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2734 of $address_data from the recipient verification was clobbered by the
2735 sender verification.
2737 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2738 was its contents. (It was OK if the option was not defined at all.)
2740 11. A "Completed" log line is now written for messages that are removed from
2741 the spool by the -Mrm option.
2743 12. $host_address is now set to the target address during the checking of
2744 ignore_target_hosts.
2746 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2747 being passed; this would have caused $sender_host_name to have been used if
2748 matching the list had actually called for a host name (not very likely,
2749 since this list is usually IP addresses). A host name is now passed as
2752 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2753 code that specifies a non-restarting handler (typically sigaction() in
2754 modern systems) in an attempt to fix a rare and obscure crash bug.
2756 15. Narrowed the window for a race in the daemon that could cause it to ignore
2757 SIGCHLD signals. This is not a major problem, because they are used only to
2758 wake it up if nothing else does.
2760 16. A malformed maildirsize file could cause Exim to calculate negative values
2761 for the mailbox size or file count. Odd effects could occur as a result.
2762 The maildirsize information is now recalculated if the size or filecount
2765 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2766 support for a long time. Removed HAVE_SYS_VFS_H.
2768 18. Updated exipick to current release from John Jetmore.
2770 19. Allow an empty sender to be matched against a lookup in an address list.
2771 Previously the only cases considered were a regular expression, or an
2774 20. Exim went into a mad DNS lookup loop when doing a callout where the
2775 host was specified on the transport, if the DNS lookup yielded more than
2778 21. The RFC2047 encoding function was originally intended for short strings
2779 such as real names; it was not keeping to the 75-character limit for
2780 encoded words that the RFC imposes. It now respects the limit, and
2781 generates multiple encoded words if necessary. To be on the safe side, I
2782 have increased the buffer size for the ${rfc2047: expansion operator from
2785 22. Failure to deliver a bounce message always caused it to be frozen, even if
2786 there was an errors_to setting on the router. The errors_to setting is now
2789 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2790 canonical form (fully expanded) before being placed in
2791 $sender_host_address.
2793 24. Updated eximstats to version 1.33
2795 25. Include certificate and key file names in error message when GnuTLS fails
2796 to set them up, because the GnuTLS error message doesn't include the name
2797 of the failing file when there is a problem reading it.
2799 26. Expand error message when OpenSSL has problems setting up cert/key files.
2802 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2803 (this can affect the format of dates).
2805 28. exim_tidydb, when checking for the continued existence of a message for
2806 which it has found a message-specific retry record, was not finding
2807 messages that were in split spool directories. Consequently, it was
2808 deleting retry records that should have stayed in existence.
2810 29. eximstats updated to version 1.35
2811 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2812 1.35 - bugfix such that pie charts by volume are generated correctly
2814 30. The SPA authentication driver was not abandoning authentication and moving
2815 on to the next authenticator when an expansion was forced to fail,
2816 contradicting the general specification for all authenticators. Instead it
2817 was generating a temporary error. It now behaves as specified.
2819 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2820 (the order specifies the preference for clients). The order is now AES256,
2821 AES128, 3DES, ARCFOUR128.
2823 31. Small patch to Sieve code - explicitly set From: when generating an
2826 32. Exim crashed if a remote delivery caused a very long error message to be
2827 recorded - for instance if somebody sent an entire SpamAssassin report back
2828 as a large number of 550 error lines. This bug was coincidentally fixed by
2829 increasing the size of one of Exim's internal buffers (big_buffer) that
2830 happened as part of the Exiscan merge. However, to be on the safe side, I
2831 have made the code more robust (and fixed the comments that describe what
2834 33. Some experimental protocols are using DNS PTR records for new purposes. The
2835 keys for these records are domain names, not reversed IP addresses. The
2836 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2837 leaves it alone. Component reversal etc. now happens only for IP addresses.
2840 34. The host_aton() function is supposed to be passed a string that is known
2841 to be a valid IP address. However, in the case of IPv6 addresses, it was
2842 not checking this. This is a hostage to fortune. Exim now panics and dies
2843 if the condition is not met. A case was found where this could be provoked
2844 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2845 components; fortuitously, this particular loophole had already been fixed
2846 by change 4.50/55 or 4.44/33 above.
2848 If there are any other similar loopholes, the new check in host_aton()
2849 itself should stop them being exploited. The report I received stated that
2850 data on the command line could provoke the exploit when Exim was running as
2851 exim, but did not say which command line option was involved. All I could
2852 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2853 running as the user.
2856 35. There was a buffer overflow vulnerability in the SPA authentication code
2857 (which came originally from the Samba project). I have added a test to the
2858 spa_base64_to_bits() function which I hope fixes it.
2861 36. The daemon start-up calls getloadavg() while still root for those OS that
2862 need the first call to be done as root, but it missed one case: when
2863 deliver_queue_load_max is set with deliver_drop_privilege. This is
2864 necessary for the benefit of the queue runner, because there is no re-exec
2865 when deliver_drop_privilege is set.
2867 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2868 was in use, was not putting the data itself into the right store pool;
2869 consequently, it could be overwritten for a subsequent message in the same
2870 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2873 38. Sometimes the final signoff response after QUIT could fail to get
2874 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2875 < 0 before doing a fflush(). This bug looks as though it goes back to the
2876 introduction of TLS in release 3.20, but "sometimes" must have been rare
2877 because the tests only now provoked it.
2883 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2884 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2885 Now it returns an integer. A number of calls were still expecting a T/F
2886 return. Fortuitously, in all cases, the tests worked in OK situations,
2887 which is the norm. However, things would have gone wrong on any write
2888 failures on the smtp file descriptor. This function is used when sending
2889 messages over SMTP and also when doing verify callouts.
2891 2. When Exim is called to do synchronous delivery of a locally submitted
2892 message (the -odf or -odi options), it no longer closes stderr before doing
2895 3. Implemented the mua_wrapper option.
2897 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2899 5. Implemented the functions header_remove(), header_testname(),
2900 header_add_at_position(), and receive_remove_recipient(), and exported them
2903 6. If an ACL "warn" statement specified the addition of headers, Exim already
2904 inserted X-ACL-Warn: at the start if there was no header name. However, it
2905 was not making this test for the second and subsequent header lines if
2906 there were newlines in the string. This meant that an invalid header could
2907 be inserted if Exim was badly configured.
2909 7. Allow an ACL "warn" statement to add header lines at the start or after all
2910 the Received: headers, as well as at the end.
2912 8. Added the rcpt_4xx retry error code.
2914 9. Added postmaster_mailfrom=xxx to callout verification option.
2916 10. Added mailfrom=xxxx to the callout verification option, for verify=
2919 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2920 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2922 12. Inserted some casts to stop certain compilers warning when using pointer
2923 differences as field lengths or precisions in printf-type calls (mostly
2924 affecting debugging statements).
2926 13. Added optional readline() support for -be (dynamically loaded).
2928 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2929 same clock tick as a message's arrival, so that its received time was the
2930 same as the "first fail" time on the retry record, and that message
2931 remained on the queue past the ultimate address timeout, every queue runner
2932 would try a delivery (because it was past the ultimate address timeout) but
2933 after another failure, the ultimate address timeout, which should have then
2934 bounced the address, did not kick in. This was a "< instead of <=" error;
2935 in most cases the first failure would have been in the next clock tick
2936 after the received time, and all would be well.
2938 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2939 being recognized when the domain list was tested by the match_domain
2940 condition in an expansion string.
2942 16. Added the ${str2b64: operator.
2944 17. Exim was always calling setrlimit() to set a large limit for the number of
2945 processes, without checking whether the existing limit was already
2946 adequate. (It did check for the limit on file descriptors.) Furthermore,
2947 errors from getrlimit() and setrlimit() were being ignored. Now they are
2948 logged to the main and panic logs, but Exim does carry on, to try to do its
2949 job under whatever limits there are.
2951 18. Imported PCRE 5.0.
2953 19. Trivial typo in log message " temporarily refused connection" (the leading
2956 20. If the log selector return_path_on_delivery was set and an address was
2957 redirected to /dev/null, the delivery process crashed because it assumed
2958 that a return path would always be set for a "successful" delivery. In this
2959 case, the whole delivery is bypassed as an optimization, and therefore no
2962 21. Internal re-arrangement: the function for sending a challenge and reading
2963 a response while authentication was assuming a zero-terminated challenge
2964 string. It's now changed to take a pointer and a length, to allow for
2965 binary data in such strings.
2967 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2969 23. Exim was not respecting finduser_retries when seeking the login of the
2970 uid under which it was called; it was always trying 10 times. (The default
2971 setting of finduser_retries is zero.) Also, it was sleeping after the final
2972 failure, which is pointless.
2974 24. Implemented tls_on_connect_ports.
2976 25. Implemented acl_smtp_predata.
2978 26. If the domain in control=submission is set empty, Exim assumes that the
2979 authenticated id is a complete email address when it generates From: or
2980 Sender: header lines.
2982 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2983 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2984 chown and chgrp in /bin and hostname in /usr/bin.
2986 28. Exim was keeping the "process log" file open after each use, just as it
2987 does for the main log. This opens the possibility of it remaining open for
2988 long periods when the USR1 signal hits a daemon. Occasional processlog
2989 errors were reported, that could have been caused by this. Anyway, it seems
2990 much more sensible not to leave this file open at all, so that is what now
2993 29. The long-running daemon process does not normally write to the log once it
2994 has entered its main loop, and it closes the log before doing so. This is
2995 so that log files can straightforwardly be renamed and moved. However,
2996 there are a couple of unusual error situations where the daemon does write
2997 log entries, and I had neglected to close the log afterwards.
2999 30. The text of an SMTP error response that was received during a remote
3000 delivery was being truncated at 512 bytes. This is too short for some of
3001 the long messages that one sometimes sees. I've increased the limit to
3004 31. It is now possible to make retry rules that apply only when a message has a
3005 specific sender, in particular, an empty sender.
3007 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3008 it possible to be selective about when SMTP synchronization is enforced.
3010 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3012 32. Implemented hosts_connection_nolog.
3014 33. Added an ACL for QUIT.
3016 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3019 35. Added mailbox_size and mailbox_filecount to appendfile.
3021 36. Added control = no_multiline_responses to ACLs.
3023 37. There was a bug in the logic of the code that waits for the clock to tick
3024 in the case where the clock went backwards by a substantial amount such
3025 that the microsecond fraction of "now" was more than the microsecond
3026 fraction of "then" (but the whole seconds number was less).
3028 38. Added support for the libradius Radius client library this is found on
3029 FreeBSD (previously only the radiusclient library was supported).
3035 1. When certain lookups returned multiple values in the form name=value, the
3036 quoting of the values was not always being done properly. Specifically:
3037 (a) If the value started with a double quote, but contained no whitespace,
3039 (b) If the value contained whitespace other than a space character (i.e.
3040 tabs or newlines or carriage returns) it was not quoted.
3041 This fix has been applied to the mysql and pgsql lookups by writing a
3042 separate quoting function and calling it from the lookup code. The fix
3043 should probably also be applied to nisplus, ibase and oracle lookups, but
3044 since I cannot test any of those, I have not disturbed their existing code.
3046 2. A hit in the callout cache for a specific address caused a log line with no
3047 reason for rejecting RCPT. Now it says "Previous (cached) callout
3048 verification failure".
3050 3. There was an off-by-one bug in the queryprogram router. An over-long
3051 return line was truncated at 256 instead of 255 characters, thereby
3052 overflowing its buffer with the terminating zero. As well as fixing this, I
3053 have increased the buffer size to 1024 (and made a note to document this).
3055 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3056 when Exim is waiting for an SMTP response from a remote server, Exim
3057 restarts its select() call on the socket, thereby resetting its timeout.
3058 This is not a problem when such interrupts are rare. Somebody set up a cron
3059 job to run exiwhat every 2 minutes, which is less than the normal select()
3060 timeout (5 or 10 minutes). This meant that the select() timeout never
3061 kicked in because it was always reset. I have fixed this by comparing the
3062 time when an interrupt arrives with the time at the start of the first call
3063 to select(). If more time than the timeout has elapsed, the interrupt is
3064 treated as a timeout.
3066 5. Some internal re-factoring in preparation for the addition of Sieve
3067 extensions (by MH). In particular, the "personal" test is moved to a
3068 separate function, and given an option for scanning Cc: and Bcc: (which is
3069 not set for Exim filters).
3071 6. When Exim created an email address using the login of the caller as the
3072 local part (e.g. when creating a From: or Sender: header line), it was not
3073 quoting the local part when it contained special characters such as @.
3075 7. Installed new OpenBSD configuration files.
3077 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3078 try to make them clearer.
3080 9. Callout options, other than the timeout value, were being ignored when
3081 verifying sender addresses in header lines. For example, when using
3083 verify = header_sender/callout=no_cache
3085 the cache was (incorrectly) being used.
3087 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3088 only the Cygwin environment.
3090 11. When return_path_on_delivery was set as a log selector, if different remote
3091 addresses in the same message used different return paths and parallel
3092 remote delivery occurred, the wrong values would sometimes be logged.
3093 (Whenever a remote delivery process finished, the return path value from
3094 the most recently started remote delivery process was logged.)
3096 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3097 lines when AUTH and/or TLS are in use. This is the "received protocol"
3098 field. Exim used to use "asmtp" for authenticated SMTP, without any
3099 indication (in the protocol name) for TLS use. Now it follows the RFC and
3100 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3101 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3102 names appear in log lines as well as in Received: header lines.
3104 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3105 extensions, and comparison tests, and to fix some bugs.
3107 14. Changes to the "personal" filter test:
3109 (1) The test was buggy in that it was just doing the equivalent of
3110 "contains" tests on header lines. For example, if a user's address was
3111 anne@some.where, the "personal" test would incorrectly be true for
3113 To: susanne@some.where
3115 This test is now done by extracting each address from the header in turn,
3116 and checking the entire address. Other tests that are part of "personal"
3117 are now done using regular expressions (for example, to check local parts
3118 of addresses in From: header lines).
3120 (2) The list of non-personal local parts in From: addresses has been
3121 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3122 taken from the Sieve specification recommendations.
3124 (3) If the message contains any header line starting with "List-" it is
3125 treated as non-personal.
3127 (4) The test for "circular" in the Subject: header line has been removed
3128 because it now seems ill-conceived.
3130 15. Minor typos in src/EDITME comments corrected.
3132 16. Installed latest exipick from John Jetmore.
3134 17. If headers_add on a router specified a text string that was too long for
3135 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3136 of string_sprintf() is now avoided.
3138 18. $message_body_size was not set (it was always zero) when running the DATA
3139 ACL and the local_scan() function.
3141 19. For the "mail" command in an Exim filter, no default was being set for
3142 the once_repeat time, causing a random time value to be used if "once" was
3143 specified. (If the value happened to be <= 0, no repeat happened.) The
3144 default is now 0s, meaning "never repeat". The "vacation" command was OK
3145 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3146 (I found it when inspecting the code).
3148 20. There is now an overall timeout for performing a callout verification. It
3149 defaults to 4 times the callout timeout, which applies to individual SMTP
3150 commands during the callout. The overall timeout applies when there is more
3151 than one host that can be tried. The timeout is checked before trying the
3152 next host. This prevents very long delays if there are a large number of
3153 hosts and all are timing out (e.g. when the network connections are timing
3154 out). The value of the overall timeout can be changed by specifying an
3155 additional sub-option for "callout", called "maxwait". For example:
3157 verify = sender/callout=5s,maxwait=20s
3159 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3160 to the end before writing, but this should make it even safer).
3162 22. Exim was forgetting that it had advertised PIPELINING for the second and
3163 subsequent messages on an SMTP connection. It was also not resetting its
3164 memory on STARTTLS and an internal HELO.
3166 23. When Exim logs an SMTP synchronization error within a session, it now
3167 records whether PIPELINING has been advertised or not.
3169 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3170 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3171 rather than long int.
3173 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3175 26. Added the never_mail option to autoreply.
3181 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3182 crash if the getsockname() call failed; this can happen if a connection is
3183 closed very soon after it is established. The problem was simply in the
3184 order in which certain operations were done, causing Exim to try to write
3185 to the SMTP stream before it had set up the file descriptor. The bug has
3186 been fixed by making things happen in the correct order.
3192 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3193 before the connection was closed, thus losing the rejection response.
3195 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3196 some early Solaris releases, but causes trouble in current releases where
3197 socklen_t is defined.
3199 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3202 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3205 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3206 that releasing the top of it at the end releases what was used for sub-
3207 expansions (unless the block got too big). However, discard this block if
3208 the first thing is a variable or header, so that we can use its block when
3209 it is dynamic (useful for very large $message_headers, for example).
3211 6. Lookups now cache *every* query, not just the most recent. A new, separate
3212 store pool is used for this. It can be recovered when all lookup caches are
3213 flushed. Lookups now release memory at the end of their result strings.
3214 This has involved some general refactoring of the lookup sources.
3216 7. Some code has been added to the store_xxx() functions to reduce the amount
3217 of flapping under certain conditions.
3219 8. log_incoming_interface used to affect only the <= reception log lines. Now
3220 it causes the local interface and port to be added to several more SMTP log
3221 lines, for example "SMTP connection from", and rejection lines.
3223 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3225 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3227 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3228 could be overwritten at the end of the current message (or the start of a
3229 new message if it was set in a HELO ACL). The value is now preserved for
3230 the duration of the SMTP connection.
3232 12. If a transport had a headers_rewrite setting, and a matching header line
3233 contained an unqualified address, that address was qualified, even if it
3234 did not match any rewriting rules. The underlying bug was that the values
3235 of the flags that permit the existence of unqualified sender and recipient
3236 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3237 non-local messages, and by -bnq for local messages) were not being
3238 preserved with the message after it was received.
3240 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3241 "next input=" as part of the text comprising the host identity instead of
3242 the correct text. The code was using the same buffer for two different
3243 strings. However, depending on which order the printing function evaluated
3244 its arguments, the bug did not always show up. Under Linux, for example, my
3245 test suite worked just fine.
3247 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3248 doesn't work with some older versions of Perl. It has been changed to "my",
3249 which in any case is probably the better facility to use.
3251 15. A really picky compiler found some instances of statements for creating
3252 error messages that either had too many or two few arguments for the format
3255 16. The size of the buffer for calls to the DNS resolver has been increased
3256 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3257 for addresses that have a lot of PTR records. This alleviates a problem; it
3258 does not fully solve it.
3260 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3261 buffer now truncates the list and logs the incident, which is the same
3262 action as happens when Exim is looking up a host name and its aliases.
3263 Previously in this situation something unpredictable would happen;
3264 sometimes it was "internal error: store_reset failed".
3266 18. If a server dropped the connection unexpectedly when an Exim client was
3267 using GnuTLS and trying to read a response, the client delivery process
3268 crashed while trying to generate an error log message.
3270 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3271 string, for example:
3273 warn message = H1: something\nH2: something
3275 the text was added as a single header line from Exim's point of view
3276 though it ended up OK in the delivered message. However, searching for the
3277 second and subsequent header lines using $h_h2: did not work. This has been
3278 fixed. Similarly, if a system filter added multiple headers in this way,
3279 the routers could not see them.
3281 20. Expanded the error message when iplsearch is called with an invalid key to
3282 suggest using net-iplsearch in a host list.
3284 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3285 ACLs are no longer actually imposed (and a message to that effect is
3288 22. If a "gecos" field in a passwd entry contained escaped characters, in
3289 particular, if it contained a \" sequence, Exim got it wrong when building
3290 a From: or a Sender: header from that name. A second bug also caused
3291 incorrect handling when an unquoted " was present following a character
3292 that needed quoting.
3294 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3295 was not being matched caselessly.
3297 24. Arranged for all hyphens in the exim.8 source to be escaped with
3300 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3301 properly. Recipient callout cache records were still being keyed to include
3302 the sender, even when use_sender was set false. This led to far more
3303 callouts that were necessary. The sender is no longer included in the key
3304 when use_sender is false.
3306 26. Added "control = submission" modifier to ACLs.
3308 27. Added the ${base62d: operator to decode base 62 numbers.
3310 28. dnsdb lookups can now access SRV records.
3312 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3313 the configuration file.
3315 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3316 (-v) mode. This makes the output for a verbose queue run more intelligible.
3318 31. Added a use_postmaster feature to recipient callouts.
3320 32. Added the $body_zerocount variable, containing the number of binary zero
3321 bytes in the message body.
3323 33. The time of last modification of the "new" subdirectory is now used as the
3324 "mailbox time last read" when there is a quota error for a maildir
3327 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3329 35. Added +ignore_unknown as a special item in host lists.
3331 36. Code for decoding IPv6 addresses in host lists is now included, even if
3332 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3333 address was recognized as an IP address, but was then not correctly decoded
3334 into binary, causing unexpected and incorrect effects when compared with
3341 1. Very minor rewording of debugging text in manualroute to say "list of
3342 hosts" instead of "hostlist".
3344 2. If verify=header_syntax was set, and a header line with an unqualified
3345 address (no domain) and a large number of spaces between the end of the
3346 name and the colon was received, the reception process suffered a buffer
3347 overflow, and (when I tested it) crashed. This was caused by some obsolete
3348 code that should have been removed. The fix is to remove it!
3350 3. When running in the test harness, delay a bit after writing a bounce
3351 message to get a bit more predictability in the log output.
3353 4. Added a call to search_tidyup() just before forking a reception process. In
3354 theory, someone could use a lookup in the expansion of smtp_accept_max_
3355 per_host which, without the tidyup, could leave open a database connection.
3357 5. Added the variables $recipient_data and $sender_data which get set from a
3358 lookup success in an ACL "recipients" or "senders" condition, or a router
3359 "senders" option, similar to $domain_data and $local_part_data.
3361 6. Moved the writing of debug_print from before to after the "senders" test
3364 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3365 problems for message scanning, either using a data ACL, or using
3366 local_scan() because the Received: header was not generated till after they
3367 were called (in order to set the time as the time of reception completion).
3368 I have revised the way this works. The header is now generated after the
3369 body is received, but before the ACL or local_scan() are called. After they
3370 are run, the timestamp in the header is updated.
3376 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3377 before starting a queue runner without re-exec. This happened only when
3378 deliver_drop_privilege was set or when the Exim user was set to root. The
3379 effect of the bug was that timeouts during subsequent deliveries caused
3380 crashes instead of being properly handled. The handler is now left at its
3381 default (and expected) setting.
3383 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3384 message, again when deliver_drop_privilege is set or Exim is run as root.
3385 The bug described in (1) was not present in this case, but the tidying up
3386 of the other signals was missing. I have made the two cases consistent.
3388 3. The ignore_target_hosts setting on a manualroute router was being ignored
3389 for hosts that were looked up using the /MX notation.
3391 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3394 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3395 operated on the sender address. After changing the $sender_address to <>
3396 for the sender address verify, Exim was re-instated it as the original
3397 (before rewriting) address, but remembering that it had rewritten it, so it
3398 wasn't rewriting it again. This bug also had the effect of breaking the
3399 sender address verification caching when the sender address was rewritten.
3401 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3402 This has been changed so that if the ip literal address matches
3403 ignore_target_hosts, the router declines.
3405 7. Added expansion conditions match_domain, match_address, and match_local_
3406 part (NOT match_host).
3408 8. The placeholder for the Received: header didn't have a length field set.
3410 9. Added code to Exim itself and to exim_lock to test for a specific race
3411 condition that could lead to file corruption when using MBX delivery. The
3412 issue is with the lockfile that is created in /tmp. If this file is removed
3413 after a process has opened it but before that process has acquired a lock,
3414 there is the potential for a second process to recreate the file and also
3415 acquire a lock. This could lead to two Exim processes writing to the file
3416 at the same time. The added code performs the same test as UW imapd; it
3417 checks after acquiring the lock that its file descriptor still refers to
3418 the same named file.
3420 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3421 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3424 11. Added the smtp_active_hostname option. If used, this will typically be made
3425 to depend on the incoming interface address. Because $interface_address is
3426 not set up until the daemon has forked a reception process, error responses
3427 that can happen earlier (such as "too many connections") no longer contain
3430 12. If an expansion in a condition on a "warn" statement fails because a lookup
3431 defers, the "warn" statement is abandoned, and the next ACL statement is
3432 processed. Previously this caused the whole ACL to be aborted.
3434 13. Added the iplsearch lookup type.
3436 14. Added ident_timeout as a log selector.
3438 15. Added tls_certificate_verified as a log selector.
3440 16. Added a global option tls_require_ciphers (compare the smtp transport
3441 option of the same name). This controls incoming TLS connections.
3443 17. I finally figured out how to make tls_require_ciphers do a similar thing
3444 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3445 before starting the TLS session.
3447 18. Tabs are now shown as \t in -bP output.
3449 19. If the log selector return_path_on_delivery was set, Exim crashed when
3450 bouncing a message because it had too many Received: header lines.
3452 20. If two routers both had headers_remove settings, and the first one included
3453 a superfluous trailing colon, the final name in the first list and the
3454 first name in the second list were incorrectly joined into one item (with a
3455 colon in the middle).
3461 1. Added -C and -D options to the exinext utility, mainly to make it easier
3462 to include in the automated testing, but these could be helpful when
3463 multiple configurations are in use.
3465 2. The exinext utility was not formatting the output nicely when there was
3466 an alternate port involved in the retry record key, nor when there was a
3467 message id as well (for retries that were specific to a specific message
3468 and a specific host). It was also confused by IPv6 addresses, because of
3469 the additional colons they contain. I have fixed the IPv4 problem, and
3470 patched it up to do a reasonable job for IPv6.
3472 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3473 delivery, the log line now contains "pipelined" if PIPELINING was used.
3475 4. An SMTP transport process used to panic and die if the bind() call to set
3476 an explicit outgoing interface failed. This has been changed; it is now
3477 treated in the same way as a connect() failure.
3479 5. A reference to $sender_host_name in the part of a conditional expansion
3480 that was being skipped was still causing a DNS lookup. This no longer
3483 6. The def: expansion condition was not recognizing references to header lines
3484 that used bh_ and bheader_.
3486 7. Added the _cache feature to named lists.
3488 8. The code for checking quota_filecount in the appendfile transport was
3489 allowing one more file than it should have been.
3491 9. For compatibility with Sendmail, the command line option
3499 and sets the incoming protocol and host name (for trusted callers). The
3500 host name and its colon can be omitted when only the protocol is to be set.
3501 Note the Exim already has two private options, -pd and -ps, that refer to
3502 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3503 "s", but I don't think that's a major issue.
3505 10. A number of refactoring changes to the code, none of which should affect
3508 (a) The number of logging options was getting close to filling up the
3509 32-bit word that was used as a bit map. I have split them into two classes:
3510 those that are passed in the argument to log_write(), and those that are
3511 only ever tested independently outside of that function. These are now in
3512 separate 32-bit words, so there is plenty of room for expansion again.
3513 There is no change in the user interface or the logging behaviour.
3515 (b) When building, for example, log lines, the code previously used a
3516 macro that called string_cat() twice, in order to add two strings. This is
3517 not really sufficiently general. Furthermore, there was one instance where
3518 it was actually wrong because one of the argument was used twice, and in
3519 one call a function was used. (As it happened, calling the function twice
3520 did not affect the overall behaviour.) The macro has been replaced by a
3521 function that can join an arbitrary number of extra strings onto a growing
3524 (c) The code for expansion conditions now uses a table and a binary chop
3525 instead of a serial search (which was left over from when there were very
3526 few conditions). Also, it now recognizes conditions like "pam" even when
3527 the relevant support is not compiled in: a suitably worded error message is
3528 given if an attempt is made to use such a condition.
3530 11. Added ${time_interval:xxxxx}.
3532 12. A bug was causing one of the ddress fields not to be passed back correctly
3533 from remote delivery subprocesses. The field in question was not being
3534 subsequently used, so this caused to problems in practice.
3536 13. Added new log selectors queue_time and deliver_time.
3538 14. Might have fixed a bug in maildirsizefile handling that threw up
3539 "unexpected character" debug warnings, and recalculated the data
3540 unnecessarily. In any case, I expanded the warning message to give more
3543 15. Added the message "Restricted characters in address" to the statements in
3544 the default ACL that block characters like @ and % in local parts.
3546 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3547 Three changes have been made:
3549 (a) There was a serious bug; a negative response to MAIL caused the whole
3550 recipient domain to be cached as invalid, thereby blocking all messages
3551 to all local parts at the same domain, from all senders. This bug has
3552 been fixed. The domain is no longer cached after a negative response to
3553 MAIL if the sender used is not empty.
3555 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3558 (c) A new callout option, "use_sender" has been added for people who want
3559 the modified behaviour.
3565 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3568 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3569 indeed breaks things for older releases.
3571 3. Added additional logging to the case where there is a problem reading data
3572 from a filter that is running in a subprocess using a pipe, in order to
3573 try to track down a specific problem.
3575 4. Testing facility fudge: when running in the test harness and attempting
3576 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3577 getting "No route to host". Convert this to a timeout.
3579 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3582 6. Some OS don't have socklen_t but use size_t instead. This affects the
3583 fifth argument of getsockopt() amongst other things. This is now
3584 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3585 can be set for individual OS. I have set it for SunOS5, OSF1, and
3586 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3587 some earlier ones do not.
3589 7. Change 4.30/15 was not doing the test caselessly.
3591 8. The standard form for an IPv6 address literal was being rejected by address
3592 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3593 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3594 this, as well as the form without the "IPv6" on the front (but only when
3595 address literals are enabled, of course).
3597 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3599 10. Exim crashed if a message with an empty sender address specified by -f
3600 encountered a router with an errors_to setting. This could be provoked only
3601 by a command such as
3605 where an empty string was supplied; "<>" did not hit this bug.
3607 11. Installed PCRE release 4.5.
3609 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3610 remained set. It is now erased.
3612 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3613 times from message ids (which are base 36 rather than the normal 62).
3615 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3616 were being counted as actual protocol errors, and logged if the log
3617 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3618 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3619 and DATA following a set of rejected RCPTs do not count as protocol errors.
3620 In other words, Exim assumes they were pipelined, though this may not
3621 actually be the case. Of course, in all cases the client gets an
3622 appropriate error code.
3624 15. If a lookup fails in an ACL condition, a message about the failure may
3625 be available; it is used if testing the ACL cannot continue, because most
3626 such messages specify what the cause of the deferral is. However, some
3627 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3628 that caused an old message to be retained and used if a later statement
3629 caused a defer, replacing the real cause of the deferral.
3631 16. If an IP address had so many PTR records that the DNS lookup buffer
3632 was not large enough to hold them, Exim could crash while trying to process
3633 the truncated data. It now detects and logs this case.
3635 17. Further to 4.21/58, another change has been made: if (and only if) the
3636 first line of a message (the first header line) ends with CRLF, a bare LF
3637 in a subsequent header line has a space inserted after it, so as not to
3638 terminate the header.
3640 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3641 unnecessarily, used atoi() instead of strtol(), and didn't check the
3642 termination when getting file sizes from file names by regex.
3644 19. Completely re-implemented the support for maildirsize files, in the light
3645 of a number of problems with the previous contributed implementation
3646 (4.30/29). In particular:
3648 . If the quota is zero, the maildirsize file is maintained, but no quota is
3651 . If the maildir directory does not exist, it is created before any attempt
3652 to write a maildirsize file.
3654 . The quota value in the file is just a cache; if the quota is changed in
3655 the transport, the new value overrides.
3657 . A regular expression is available for excluding directories from the
3660 20. The autoreply transport checks the characters in options that define the
3661 message's headers; it allows continued headers, but it was checking with
3662 isspace() after an embedded newline instead of explicitly looking for a
3665 21. If all the "regular" hosts to which an address was routed had passed their
3666 expiry times, and had not reached their retry times, the address was
3667 bounced, even if fallback hosts were defined. Now Exim should go on to try
3670 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3671 equivalent code in the SMTP transport. Some hosts send humungous responses
3672 to HELO/EHLO, more than 1024 it seems.
3674 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3675 gives compiler warnings in some environments. I've now done it "properly",
3678 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3679 (because of problems with the built-in one) was declared to return uschar *
3680 instead of char *, causing compiler failure.
3682 25. Fixed a file descriptor leak when processing alias/forward files.
3684 26. Fixed a minor format string issue in dbfn.c.
3686 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3688 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3689 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3692 29. When a sender address is verified, it is cached, to save repeating the test
3693 when there is more than one recipient in a message. However, when the
3694 verification involves a callout, it is possible for different callout
3695 options to be set for different recipients. It is too complicated to keep
3696 track of this in the cache, so now Exim always runs a verification when a
3697 callout is required, relying on the callout cache for the optimization.
3698 The overhead is duplication of the address routing, but this should not be
3701 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3702 to be verified with callout=postmaster, and the main callout worked but the
3703 postmaster check failed, the verification correctly failed. However, if a
3704 subsequent RCPT command asked for sender verification *without* the
3705 postmaster check, incorrect caching caused this verification also to fail,
3708 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3709 it was not caching the DNS options (qualify_single, search_parents) that
3710 were used when the lookup failed. A subsequent lookup with different
3711 options therefore always gave the same answer, though there were cases
3712 where it should not have. (Example: a "domains = !$mx_any" option on a
3713 dnslookup router: the "domains" option is always processed without any
3714 widening, but the router might have qualify_single set.) Now Exim uses the
3715 cached value only when the same options are set.
3717 32. Added John Jetmore's "exipick" utility to the distribution.
3719 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3720 than a timeout (e.g. a certificate is required, and is not provided), an
3721 Exim server now closes the connection immediately. Previously it waited for
3722 the client to close - but if the client is SSL, it seems that they each
3723 wait for each other, leading to a delay before one of them times out.
3725 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3726 maintained 0.8.x compatibility because I don't think many are using it, and
3727 it is clearly obsolete.
3729 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3732 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3733 client certificate was expired. A simple patch fixes this, though I don't
3734 understand the full logic of why the verify callback is called multiple
3737 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3738 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3739 which causes problems with some clients (such as the Certicom SSL Plus
3740 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3741 disables the coutermeasure allowing Eudora to connect."
3743 38. Exim was not checking that a write() to a log file succeeded. This could
3744 lead to Bad Things if a log got too big, in particular if it hit a file
3745 size limit. Exim now panics and dies if it cannot write to a log file, just
3746 as it does if it cannot open a log file.
3748 39. Modified OS/Makefile-Linux so that it now contains
3750 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3752 The two -D definitions ensure that Exim is compiled with large file
3753 support, which makes it possible to handle log files that are bigger than
3756 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3757 instance) a domain was checked against a named list that involved a lookup,
3758 causing $domain_data to be set, then another domain was checked against the
3759 same list, then the first domain was re-checked, the value of $domain_data
3760 after the final check could be wrong. In particular, if the second check
3761 failed, it could be set empty. This bug probably also applied to
3764 41. The strip_trailing_dot option was not being applied to the address given
3765 with the -f command-line option.
3767 42. The code for reading a message's header from the spool was incrementing
3768 $received_count, but never initializing it. This meant that the value was
3769 incorrect (doubled) while delivering a message in the same process in which
3770 it was received. In the most common configuration of Exim, this never
3771 happens - a fresh exec is done - but it can happen when
3772 deliver_drop_privilege is set.
3774 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3775 it now includes up to 150 characters of the unexpected data in the log
3778 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3779 and building data strings. The size of both of these buffers was 10 000
3780 bytes - far larger than anybody would *ever* want, thought I. Needless to
3781 say, somebody hit the limit. I have increased the maximum line length to
3782 20 000 and the maximum data length of concatenated lines to 100 000. I have
3783 also fixed two bugs, because there was no checking on these buffers. Tsk,
3784 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3785 buffer is too small.
3787 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3788 lsearch lookups. Now it does.
3790 46. When parsing a route_list item in a manualroute router, a fixed-length
3791 buffer was used for the list of hosts. I made this 1024 bytes long,
3792 thinking that nobody would ever have a list of hosts that long. Wrong.
3793 Somebody had a whole pile of complicated expansion conditions, and the
3794 string was silently truncated, leading to an expansion error. It turns out
3795 that it is easier to change to an unlimited length (owing to other changes
3796 that have happened since this code was originally written) than to build
3797 structure for giving a limitation error. The length of the item that
3798 expands into the list of hosts is now unlimited.
3800 47. The lsearch lookup could not handle data where the length of text line was
3801 more than 4095 characters. Such lines were truncated, leading to shortened
3802 data being returned. It should now handle lines of any length.
3804 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3805 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3808 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3809 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3810 no longer happens in comment lines. A list of replacements is now placed
3811 at the head of all of the source files, except those whose only change is
3812 to replace PERL_COMMAND in the very first #! line.
3814 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3815 messages on the queue, with a bottom-up merge sort, using code contributed
3816 by Michael Haardt. This should make operations like -bp somewhat faster on
3817 large queues. It won't affect queue runners, except when queue_run_in_order
3820 51. Installed eximstats 1.31 in the distribution.
3822 52. Added support for SRV lookups to the dnslookup router.
3824 53. If an ACL referred to $message_body or $message_body_end, the value was not
3825 reset for any messages that followed in the same SMTP session.
3827 54. The store-handling optimization for building very long strings was not
3828 differentiating between the different store pools. I don't think this
3829 actually made any difference in practice, but I've tidied it.
3831 55. While running the routers to verify a sender address, $sender_address
3832 was still set to the sender address. This is wrong, because when routing to
3833 send a bounce to the sender, it would be empty. Therefore, I have changed
3834 it so that, while verifying a sender address, $sender_address is set to <>.
3835 (There is no change to what happens when verifying a recipient address.)
3837 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3838 target A or AAAA records (if not already returned) without resetting the
3839 qualify_single or search_parents options of the DNS resolver. These are
3840 inappropriate in this case because the targets of MX and SRV records must
3841 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3842 target that, when qualified, matched something in the local domain. These
3843 two options are now turned off when doing these lookups.
3845 57. It seems that at least some releases of Reiserfs (which does not have the
3846 concept of a fixed number of inodes) returns zero and not -1 for the
3847 number of available inodes. This interacted badly with check_spool_inodes,
3848 which assumed that -1 was the "no such thing" setting. What I have done is
3849 to check that the total number of inodes is greater than zero before doing
3850 the test of how many are available.
3852 58. When a "warn" ACL statement has a log_message modifier, the message is
3853 remembered, and not repeated. This is to avoid a lot of repetition when a
3854 message has many recipients that cause the same warning to be written.
3855 Howewer, Exim was preserving the list of already written lines for an
3856 entire SMTP session, which doesn't seem right. The memory is now reset if a
3857 new message is started.
3859 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3860 debugging output unless log_rewrite was also set.
3862 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3863 of (int)(handle) when we know that handle contains (void *)(-1).
3865 61. The Exim daemon panic-logs an error return when it closes the incoming
3866 connection. However "connection reset by peer" seems to be common, and
3867 isn't really an error worthy of noting specially, so that particular error
3870 62. When Exim is trying to find all the local interfaces, it used to panic and
3871 die if the ioctl to get the interface flags failed. However, it seems that
3872 on at least one OS (Solaris 9) it is possible to have an interface that is
3873 included in the list of interfaces, but for which you get a failure error
3874 for this call. This happens when the interface is not "plumbed" into a
3875 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3876 failure of the "get flags" call assumes that the interface is down.
3878 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3879 makes life easier for people who are doing arithmetic on fields extracted
3880 from dates, where you often get leading zeros that should not be
3881 interpreted as octal.
3883 64. Added qualify_domain to the redirect router, to override the global
3886 65. If a pathologically long header line contained very many addresses (the
3887 report of this problem mentioned 10 000) and each of them was rewritten,
3888 Exim could use up a very large amount of memory. (It kept on making new
3889 copies of the header line as it rewrote, and never released the old ones.)
3890 At the expense of a bit more processing, the header rewriting function has
3891 been changed so that it no longer eats memory in this way.
3893 66. The generation of the Received: header has been moved from the time that a
3894 message starts to be received, to the time that it finishes. The timestamp
3895 in the Received: header should now be very close to that of the <= log
3896 line. There are two side-effects of this change:
3898 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3899 logged header lines no longer include the local Received: line, because
3900 it has not yet been created. The same applies to a copy of the message
3901 that is returned to a non-SMTP sender when a message is rejected.
3903 (b) When a filter file is tested using -bf, no additional Received: header
3904 is added to the test message. After some thought, I decided that this
3907 This change does not affect the value of $received_for. It is still set
3908 after address rewriting, but before local_scan() is called.
3910 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3912 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3913 gave an unhelpful panic error message, and a defer error. I have managed to
3914 change this behaviour so that it now rejects any supplied certificate,
3915 which seems right, as the list of acceptable certificates is empty.
3917 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3918 gave an unhelpful defer error. I have not managed to make this reject any
3919 supplied certificates, but the error message it gives is "no certificate
3920 supplied", which is not helpful.
3922 70. exigrep's output now also includes lines that are not associated with any
3923 message, but which match the given pattern. Implemented by a patch from
3924 Martin Sluka, which also tidied up the Perl a bit.
3926 71. Recipient callout verification, like sender verification, was using <> in
3927 the MAIL FROM command. This isn't really the right thing, since the actual
3928 sender may affect whether the remote host accepts the recipient or not. I
3929 have changed it to use the actual sender in the callout; this means that
3930 the cache record is now keyed on a recipient/sender pair, not just the
3931 recipient address. There doesn't seem to be a real danger of callout loops,
3932 since a callout by the remote host to check the sender would use <>.
3933 [SEE ABOVE: changed after hitting problems.]
3935 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3936 temporary errors. However, in the case of such a code being given after
3937 the end of a data transmission (i.e. after ".") Exim was failing to write
3938 a retry record for the message. (Yes, there was some broken host that was
3939 actually sending 8xx at this point.)
3941 73. An unknown lookup type in a host list could cause Exim to panic-die when
3942 the list was checked. (An example that provoked this was putting <; in the
3943 middle of a list instead of at the start.) If this happened during a DATA
3944 ACL check, a -D file could be left lying around. This kind of configuration
3945 error no longer causes Exim to die; instead it causes a defer errror. The
3946 incident is still logged to the main and panic logs.
3948 74. Buglet left over from Exim 3 conversion. The message "too many messages
3949 in one connection" was written to the rejectlog but not the mainlog, except
3950 when address rewriting (yes!) was being logged.
3952 75. Added write_rejectlog option.
3954 76. When a system filter was run not as root (that is, when system_filter_user
3955 was set), the values of the $n variables were not being returned to the
3956 main process; thus, they were not subsequently available in the $sn
3959 77. Added +return_path_on_delivery log selector.
3961 78. A connection timeout was being treated differently from recipients deferred
3962 when testing hosts_max_try with a message that was older than the host's
3963 retry timeout. (The host should not be counted, thus allowing all hosts to
3964 be tried at least once before bouncing.) This may have been the cause of an
3965 occasionally reported bug whereby a message would remain on the queue
3966 longer than the retry timeout, but would be bounced if a delivery was
3967 forced. I say "may" because I never totally pinned down the problem;
3968 setting up timeout/retry tests is difficult. See also the next item.
3970 79. The ultimate address timeout was not being applied to errors that involved
3971 a combination of host plus message (for example, a timeout on a MAIL
3972 command). When an address resolved to a number of possible hosts, and they
3973 were not all tried for each delivery (e.g. because of hosts_max_try), a
3974 message could remain on the queue longer than the retry timeout.
3976 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3979 81. Fixed an obscure SMTP outgoing bug which required at least the following
3980 conditions: (a) there was another message waiting for the same server;
3981 (b) the server returned 5xx to all RCPT commands in the first message so
3982 that the message was not completed; (c) the server dropped the connection
3983 or gave a negative response to the RSET that Exim sends to abort the
3984 transaction. The observed case was a dropped connection after DATA that had
3985 been sent in pipelining mode. That is, the server had advertised PIPELINING
3986 but was not implementing it correctly. The effect of the bug was incorrect
3987 behaviour, such as trying another host, and this could lead to a crash.
3993 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3994 and daemon.c were passed as pointers to ints; they should have been
3995 pointers to socklen_t variables (which are typically unsigned ints).
3997 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4000 3. Fixed a really odd bug that affected only the testing scheme; patching a
4001 certain fixed string in the binary changed the value of another string that
4002 happened to be identical to the end of the original first string.
4004 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4005 name", it returns that address as the IP address. On some operating
4006 systems (e.g. Solaris), it also passes back the IP address string as the
4007 "host name". However, on others (e.g. Linux), it passes back an empty
4008 string. Exim wasn't checking for this, and was changing the host name to an
4009 empty string, assuming it had been canonicized.
4011 5. Although rare, it is permitted to have more than one PTR record for a given
4012 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4013 all the names associated with an address, because they do in Solaris.
4014 However, it seems that they do not in Linux for data that comes from the
4015 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4016 I found this out when I moved to a new Linux workstation and tried to run
4017 the Exim test suite.
4019 To get round this problem I have changed the code so that it now does its
4020 own call to the DNS to look up PTR records when searching for a host name.
4021 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4022 addresses that are only in /etc/hosts are still found.
4024 This behaviour is, however, controlled by an option called host_lookup_
4025 order, which defaults to "bydns:byaddr". If people want to use the other
4026 order, or indeed, just use one or the other means of lookup, they can
4027 specify it in this variable.
4029 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4030 some operating systems, this comes back from gethostbyaddr() as an empty
4031 string, and this is what Exim used to test for. However, it seems that in
4032 other systems, "." is yielded. Exim now tests for this case too.
4034 7. The values of check_spool_space and check_log_space are now held internally
4035 as a number of kilobytes instead of an absolute number of bytes. If a
4036 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4037 kilobyte. This means that much larger values can be stored.
4039 8. Exim monitor: an attempt to get the action menu when not actually pointing
4040 at a message produces an empty menu entitled "No message selected". This
4041 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4042 no entries in it ("Shell widget menu has zero width and/or height"). So I
4043 have added a single, blank menu entry in this case.
4045 9. Added ${quote_local_part.
4047 10. MIME decoding is now applied to the contents of Subject: header lines when
4050 11. Now that a reference to $sender_host_address automatically causes a reverse
4051 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4052 host lookup before query-style lookups in lists that might use this
4053 variable. This has therefore been abolished, and the "net-" prefix is no
4054 longer necessary for query-style lookups.
4056 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4057 have been a typo for LFLAGS, so it has been changed.
4059 13. The install script calls Exim with "-C /dev/null" in order to find the
4060 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4061 to be output. Howeve, since Exim outputs its version number before the
4062 error, it didn't break the script. It just looked ugly. I fixed this by
4063 always allowing "-C /dev/null" if the caller is root.
4065 14. Ignore overlarge ACL variable number when reading spool file - insurance
4066 against a later release with more variables having written the file.
4068 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4069 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4070 this, as well as the form without the "IPv6" on the front.
4072 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4073 OS/Makefile-Darwin file.
4075 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4076 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4078 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4079 variable. However, some OS use ints for inodes. I've added cast to long int
4080 to get rid of the compiler warning.
4082 19. I had forgotten to lock out "/../" in configuration file names when
4083 ALT_CONFIG_PREFIX was set.
4085 20. Routers used for verification do not need to specify transports. However,
4086 if such a router generated a host list, and callout was configured, Exim
4087 crashed, because it could not find a port number from the (non-existent)
4088 transport. It now assumes port 25 in this circumstance.
4090 21. Added the -t option to exigrep.
4092 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4093 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4094 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4095 them is set, LOOKUP_LSEARCH is forced.
4097 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4099 24. Added sender and host information to the "rejected by local_scan()" log
4100 line; previously there was no indication of these.
4102 25. Added .include_if_exists.
4104 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4105 Linux. It turns out that not all file systems support this. Apparently some
4106 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4107 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4108 not supported on the file descriptor, is now ignored when Exim is trying to
4109 sync a directory. This applies only to Linux.
4111 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4113 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4114 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4117 29. Added support for maildirsize files from supplied patch (modified a bit).
4119 30. The use of :fail: followed by an empty string could lead Exim to respond to
4120 sender verification failures with (e.g.):
4122 550 Verification failed for <xxx>
4123 550 Sender verify failed
4125 where the first response line was missing the '-' that indicates it is not
4126 the final line of the response.
4128 31. The loop for finding the name of the user that called Exim had a hardwired
4129 limit of 10; it now uses the value of finduser_retries, which is used for
4130 all other user lookups.
4132 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4135 33. Exim was neglecting to zero errno before one call of strtol() when
4136 expanding a string and expecting an integer value. On some systems this
4137 resulted in spurious "integer overflow" errors. Also, it was casting the
4138 result into an int without checking.
4140 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4141 did not work. The code looks as if it has *never* worked, though it appears
4142 to have been documented since at least releast 1.62. I have made it work.
4144 35. The "timeout_DNS" error in retry rules, also documented since at least
4145 1.62, also never worked. As it isn't clear exactly what this means, and
4146 clearly it isn't a major issue, I have abolished the feature by treating it
4147 as "timeout", and writing a warning to the main and panic logs.
4149 36. The display of retry rules for -brt wasn't always showing the error code
4152 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4153 timeout_connect_A, timeout_connect_MX.
4155 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4156 to the empty sender.
4158 39. The daemon was not analysing the content of -oX till after it had closed
4159 stderr and disconnected from the controlling terminal. This meant that any
4160 syntax errors were only noted on the panic log, and the return code from
4161 the command was 0. By re-arranging the code a little, I've made the
4162 decoding happen first, so such errors now appear on stderr, and the return
4163 code is 1. However, the actual setting up of the sockets still happens in
4164 the disconnected process, so errors there are still only recorded on the
4167 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4168 connections (as happens on some IP stacks) was logged at start up time as
4169 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4170 it from "IPv6 and IPv4", which means that two separate sockets are being
4173 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4174 says whether AF_INET or AF_INET6 was passed as an argument.
4176 42. Exiwhat output was messed up when time zones were included in log
4179 43. Exiwhat now gives more information about the daemon's listening ports,
4180 and whether -tls-on-connect was used.
4182 44. The "port" option of the smtp transport is now expanded.
4184 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4185 silently ignored. Now an error message is written to the main and panic
4188 46. There's a new ACL modifier called "logwrite" which writes to a log file
4189 as soon as it is encountered.
4191 47. Added $local_user_uid and $local_user_gid at routing time.
4193 48. Exim crashed when trying to verify a sender address that was being
4196 49. Exim was recognizing only a space character after ".include". It now also
4197 recognizes a tab character.
4199 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4200 extracting the relevant information from the specification. The man page no
4201 longer contains scrambled data for the -d option, and I've added a section
4202 at the front about calling Exim under different names.
4204 51. Added "extra_headers" argument to the "mail" command in filter files.
4206 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4209 53. Installed eximstats 1.29.
4211 54. Added transport_filter_timeout as a generic transport option.
4213 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4214 Cc: header lines. This was required by RFC 822, but it not required by RFC
4217 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4218 incoming messages that did not have them. Now it does so only if the
4219 message originates locally, that is, if there is no associated remote host
4220 address. When Resent- header lines are present, this applies to the Resent-
4221 lines rather than the non-Resent- lines.
4223 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4224 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4226 58. Messages for configuration errors now include the name of the main
4227 configuration file - useful now that there may be more than one file in a
4228 list (.included file names were always shown).
4230 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4231 for those rare installations that do not start the daemon as root or run it
4232 setuid root. I've cut out the call to initgroups() if the daemon is not
4235 60. The Exim user and group can now be bound into the binary as text strings
4236 that are looked up at the start of Exim's processing.
4238 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4240 62. Added $mailstore_basename variable.
4242 63. Installed patch to sieve.c from Michael Haardt.
4244 64. When Exim failed to open the panic log after failing to open the main log,
4245 the original message it was trying to log was written to stderr and debug
4246 output, but if they were not available (the usual case in production), it
4247 was lost. Now it is written to syslog before the two lines that record the
4248 failures to open the logs.
4250 65. Users' Exim filters run in subprocesses under the user's uid. It is
4251 possible for a "deliver" command or an alias in a "personal" command to
4252 provoke an address rewrite. If logging of address rewriting is configured,
4253 this fails because the process is not running as root or exim. There may be
4254 a better way of dealing with this, but for the moment (because 4.30 needs
4255 to be released), I have disabled address rewrite logging when running a
4256 filter in a non-root, non-exim process.
4262 1. The buildconfig auxiliary program wasn't quoting the value set for
4263 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4264 not defined. This bug was masked in 4.22 by the effect that was fixed in
4267 2. Some messages that were rejected after a message id was allocated were
4268 shown as "incomplete" by exigrep. It no longer does this for messages that
4269 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4271 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4272 have allow_domain_literals set, the ID did not get logged in the <= line.
4273 Domain literals are now always recognized in Message-ID: header lines.
4275 4. The first argument for a ${extract expansion item is the key name or field
4276 number. Leading and trailing spaces in this item were not being ignored,
4277 causing some misleading effects.
4279 5. When deliver_drop_privilege was set, single queue runner processes started
4280 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4281 same command in the process it spins off) were not dropping privilege.
4283 6. When the daemon running as "exim" started a queue runner, it always
4284 re-executed Exim in the spun-off process. This is a waste of effort when
4285 deliver_drop_privilege is set. The new process now just calls the
4286 queue-runner function directly.
4292 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4295 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4296 ignored. Though the use of -oP was forcing the writing of a pid file, it
4297 was always written to the default place.
4299 3. If the message "no IP address found for host xxxx" is generated during
4300 incoming verification, it is now followed by identification of the incoming
4301 connection (so you can more easily find what provoked it).
4303 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4305 5. Added some features to "harden" Exim a bit more against certain attacks:
4307 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4308 be put in Local/Makefile. This is like the never_users runtime option,
4309 but it cannot be overridden. The default setting is "root".
4311 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4312 prefix string with which any file named in a -C command line option
4315 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4316 is retained for -C and -D only if the caller of Exim is root. Without
4317 it, the exim user may also use -C and -D and retain privilege.
4319 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4320 command line option is disabled.
4322 6. Macro names set by the -D option must start with an upper case letter, just
4323 like macro names defined in the configuration file.
4325 7. Added "dereference=" facility to LDAP.
4327 8. Two instances of the typo "uknown" in the source files are fixed.
4329 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4330 the Configure-Makefile script screwed up while processing it.
4332 10. Incorporated PCRE 4.4.
4334 11. The SMTP synchronization check was not operating right at the start of an
4335 SMTP session. For example, it could not catch a HELO sent before the client
4336 waited for the greeting. There is now a check for outstanding input at the
4337 point when the greeting is written. Because of the duplex, asynchronous
4338 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4339 way, but not yet received, when the check is performed.
4341 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4342 on TCP/IP sockets, because this apparently causes some broken clients to
4345 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4346 unchanged) from the Cygwin maintainer.
4348 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4349 was supported instead of testing for mbx. Effectively a typo.
4351 15. The spa authenticator server code was not checking that the input it
4352 received was valid base64.
4354 16. The debug output line for the "set" modifier in ACLs was not showing the
4355 name of the variable that was being set.
4357 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4359 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4360 lookup on its own. Something else was needed to trigger the lookup. For
4361 example, a match in host_lookup or the need for a host name in a host list.
4362 Now, if $sender_host_name is referenced and the host name has not yet been
4363 looked up, a lookup is performed. If the lookup fails, the variable remains
4364 empty, and $host_lookup_failed is set to "1".
4366 19. Added "eqi" as a case-independent comparison operator.
4368 20. The saslauthd authentication condition could segfault if neither service
4369 nor realm was specified.
4371 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4372 error message that was logged was misleading, and incoming SMTP
4373 connections were dropped. The message is now more accurate, and temporary
4374 errors are given to SMTP connections.
4376 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4377 (except RSET) with a 421 error, until QUIT is received. However, it was
4378 failing to send a response to QUIT.
4380 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4381 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4382 ACL is now run after the helo_try_verify_hosts code.
4384 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4385 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4386 case-independent, so other case variants are also recognized). Apparently
4387 some systems use these upper case variants.
4389 25. If more than two messages were waiting for the same host, and a transport
4390 filter was specified for the transport, Exim sent two messages over the
4391 same TCP/IP connection, and then failed with "socket operation on non-
4392 socket" when it tried to send the third.
4394 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4396 27. The extern definition of crypt16() in expand.c was not being excluded when
4397 the OS had its own crypt16() function.
4399 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4400 as a preferred synonym for return_size_limit, both as an option and as an
4403 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4405 30. Changed the default configuration ACL to relax the local part checking rule
4406 for addresses that are not in any local domains. For these addresses,
4407 slashes and pipe symbols are allowed within local parts, but the sequence
4408 /../ is explicitly forbidden.
4410 31. SPA server authentication was not clearing the challenge buffer before
4413 32. log_message in a "warn" ACL statement was writing to the reject log as
4414 well as to the main log, which contradicts the documentation and doesn't
4415 seem right (because no rejection is happening). So I have stopped it.
4417 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4418 However, I am unable to do any testing of this.
4420 34. Fixed an infelicity in the appendfile transport. When checking directories
4421 for a mailbox, to see if any needed to be created, it was accidentally
4422 using path names with one or more superfluous leading slashes; tracing
4423 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4425 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4426 contents are added to the log line that is written for every discarded
4427 recipient. (Previously a log_message setting was ignored.)
4429 36. The ${quote: operator now quotes the string if it is empty.
4431 37. The install script runs exim in order to find its version number. If for
4432 some reason other than non-existence or emptiness, which it checks, it
4433 could not run './exim', it was installing it with an empty version number,
4434 i.e. as "exim-". This error state is now caught, and the installation is
4437 38. An argument was missing from the function that creates an error message
4438 when Exim fails to connect to the socket for saslauthd authentication.
4439 This could cause Exim to crash, or give a corrupted message.
4441 39. Added isip, isip4, and isip6 to ${if conditions.
4443 40. The ACL variables $acl_xx are now saved with the message, and can be
4444 accessed later in routers, transports, and filters.
4446 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4447 strings in the file are not string-expanded.
4449 42. If a MAIL command specified a SIZE value that was too large to fit into an
4450 int variable, the check against message_size_limit failed. Such values are
4451 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4452 day this will have to be increased, but I don't think I want to be around
4453 when emails are that large.
4460 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4461 iconv() is not standard in FreeBSD.
4463 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4464 IPv6 enabled. The observed symptom was a segmentation fault on return from
4465 the function os_common_find_running_interfaces() in src/os.c.
4467 3. In the check_special_case() function in daemon.c I had used "errno" as an
4468 argument name, which causes warnings on some systems. This was basically a
4469 typo, since it was named "eno" in the comments!
4471 4. The code that waits for the clock to tick (at a resolution of some fraction
4472 of a second) so as to ensure message-id uniqueness was always waiting for
4473 at least one whole tick, when it could have waited for less. [This is
4474 almost certainly not relevant at current processor speeds, where it is
4475 unlikely to ever wait at all. But we try to future-proof.]
4477 5. The function that sleeps for a time interval that includes fractions of a
4478 second contained a race. It did not block SIGALRM between setting the
4479 timer, and suspending (a couple of lines later). If the interval was short
4480 and the sigsuspend() was delayed until after it had expired, the suspension
4481 never ended. On busy systems this could lead to processes getting stuck for
4484 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4485 process, before it forks any delivery processes. The open lookup caching
4486 mechanism meant that the open file or database connection was passed into
4487 the delivery process. The problem was that delivery processes always tidy
4488 up cached lookup data. This could cause a problem for the next delivery
4489 process started by the queue runner, because the external queue runner
4490 process does not know about the closure. So the next delivery process
4491 still has data in the lookup cache. In the case of a file lookup, there was
4492 no problem because closing a file descriptor in a subprocess doesn't affect
4493 the parent. However, if the lookup was caching a connection to a database,
4494 the connection was closed, and the second delivery process was likely to
4495 see errors such as "PGSQL: query failed: server closed the connection
4496 unexpectedly". The problem has been fixed by closing all cached lookups
4497 in a queue runner before running a delivery process.
4499 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4500 seem to have the "const" qualifier which it has on other OS. I've
4503 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4504 *of the same type* (client or server) with the same public name that an
4505 error should be diagnosed.
4507 9. When Exim looked up a host name for an IP address, but failed to find the
4508 original IP address when looking up the host name (a safety check), it
4509 output the message "<ip address> does not match any IP for NULL", which was
4510 confusing, to say the least. The bug was that the host name should have
4511 appeared instead of "NULL".
4513 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4514 user that is built into the binary, and the -C or -D options is used, root
4515 privilege is dropped before the configuration file is read. In addition,
4516 logging is switched to stderr instead of the normal log files. If the
4517 configuration then re-defines the Exim user, the unprivileged environment
4518 is probably not what is expected, so Exim logs a panic warning message (but
4521 However, if deliver_drop_privilege is set, the unprivileged state may well
4522 be exactly what is intended, so the warning has been cut out in that case,
4523 and Exim is allowed to try to write to its normal log files.
4529 1. smtp_return_error_details was not giving details for temporary sender
4530 or receiver verification errors.
4532 2. Diagnose a configuration error if two authenticators have the same public
4535 3. Exim used not to create the message log file for a message until the first
4536 delivery attempt. This could be confusing when incoming messages were held
4537 for policy or load reasons. The message log file is now created at the time
4538 the message is received, and an initial "Received" line is written to it.
4540 4. The automatically generated man page for command line options had a minor
4541 bug that caused no ill effects; however, a more serious problem was that
4542 the procedure for building the man page automatically didn't always
4543 operate. Consequently, release 4.20 contains an out-of-date version. This
4544 shouldn't happen again.
4546 5. When building Exim with embedded Perl support, the script that builds the
4547 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4548 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4550 6. The freeze_tell option was not being used for messages that were frozen on
4551 arrival, either by an ACL or by local_scan().
4553 7. Added the smtp_incomplete_transaction log selector.
4555 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4556 was accepting AUTH without a new EHLO.
4558 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4559 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4562 10. Logging of TCP/IP connections (when configured) now happens in the main
4563 daemon process instead of the child process, so that the TCP/IP connection
4564 count is more accurate (but it can never be perfect).
4566 11. The use of "drop" in a nested ACL was not being handled correctly in the
4567 outer ACL. Now, if condition failure induced by the nested "drop" causes
4568 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4569 or "require"), the connection is dropped.
4571 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4572 that yield "discard" can now be used with an "accept" or a "discard" verb,
4573 but an error is generated for any others (because I can't see a useful way
4574 to define what should happen).
4576 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4577 are now processed in the same way as lines in the Exim configuration file.
4578 In particular, continuation lines are supported.
4580 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4582 15. Added -ti meaning -t -i.
4584 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4585 domains, and warn by logging if others are found.
4587 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4588 returned from GIFCONF when Exim is trying to find the list of interfaces on
4589 a host. The code in os.c has been modified to copy each ifreq to an aligned
4590 structure in all cases.
4592 Also, in some cases, the returned ifreq's were being copied to a 'struct
4593 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4594 means the last couple of bytes of an IPv6 address could be chopped if the
4595 ifreq contained only a normal sockaddr (14 bytes storage).
4597 18. Named domain lists were not supported in the hosts_treat_as_local option.
4598 An entry such as +xxxx was not recognized, and was treated as a literal
4601 19. Ensure that header lines added by a DATA ACL are included in the reject log
4602 if the ACL subsequently rejects the message.
4604 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4605 MD5 (which is deprecated).
4607 21. When testing a filter file using -bf, Exim was writing a message when it
4608 took the sender from a "From " line in the message, but it was not doing so
4609 when it took $return_path from a Return-Path: header line. It now does.
4611 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4612 with a valid header line field name (a series of printing characters
4613 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4615 23. Changed "disc" in the source to "disk" to conform to the documentation and
4616 the book and for uniformity.
4618 24. Ignore Sendmail's -Ooption=value command line item.
4620 25. When execve() failed while trying to run a command in a pipe transport,
4621 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4622 could be confused with a return value of 69 from the command itself. This
4623 has been changed to 127, the value the shell returns if it is asked to run
4624 a non-existent command. The wording for the related log line suggests a
4625 non-existent command as the problem.
4627 26. If received_header_text expands to an empty string, do not add a Received:
4628 header line to the message. (Well, it adds a token one on the spool, but
4629 marks it "old" so that it doesn't get used or transmitted.)
4631 27. Installed eximstats 1.28 (addition of -nt option).
4633 28. There was no check for failure on the call to getsockname() in the daemon
4634 code. This can fail if there is a shortage of resources on the system, with
4635 ENOMEM, for example. A temporary error is now given on failure.
4637 29. Contrary to the C standard, it seems that in some environments, the
4638 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4639 program. Exim now does this explicitly; it affects the formatting of
4640 timestamps using strftime().
4642 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4643 complaints. I've now initialized all the variables, to avoid this.
4645 32. Header lines added by a system filter were not being "seen" during
4646 transport-time rewrites.
4648 33. The info_callback() function passed to OpenSSL is set up with type void
4649 (*)(SSL *, int, int), as described somewhere. However, when calling the
4650 function (actually a macro) that sets it up, the type void(*)() is
4651 expected. I've put in a cast to prevent warnings from picky compilers.
4653 34. If a DNS black list lookup found a CNAME record, but there were no A
4654 records associated with the domain it pointed at, Exim crashed.
4656 35. If a DNS black list lookup returned more than one A record, Exim ignored
4657 all but the first. It now scans all returned addresses if a particular IP
4658 value is being sought. In this situation, the contents of the
4659 $dnslist_value variable are a list of all the addresses, separated by a
4662 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4663 to accept a host name and all its aliases if the forward lookup for any of
4664 them yielded the IP address of the incoming connection. Now it accepts only
4665 those names whose forward lookup yields the correct IP address. Any other
4666 names are discarded. This closes a loophole whereby a rogue DNS
4667 administrator could create reverse DNS records to break through a
4668 wildcarded host restriction in an ACL.
4670 37. If a user filter or a system filter that ran in a subprocess used any of
4671 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4672 the wrong values were passed to the pipe command ($thisaddress had the
4673 value of $0, $0 had the value of $1, etc). This bug was introduced by
4674 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4676 38. Improved the line breaking for long SMTP error messages from ACLs.
4677 Previously, if there was no break point between 40 and 75 characters, Exim
4678 left the rest of the message alone. Two changes have been made: (a) I've
4679 reduced the minimum length to 35 characters; (b) if it can't find a break
4680 point between 35 and 75 characters, it looks ahead and uses the first one
4681 that it finds. This may give the occasional overlong line, but at least the
4682 remaining text gets split now.
4684 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4685 file descriptors might be low, and that setting 1000 would always raise it.
4686 It turns out that in some environments, the limit is already over 1000 and
4687 that lowering it causes trouble. So now Exim takes care not to decrease it.
4689 40. When delivering a message, the value of $return_path is set to $sender_
4690 address at the start of routing (routers may change the value). By an
4691 oversight, this default was not being set up when an address was tested by
4692 -bt or -bv, which affected the outcome if any router or filter referred to
4695 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4696 log only when "message" or "log_message" are set. However, if one of the
4697 conditions was an address verification, or a call to a nested ACL, the
4698 messages generated by the underlying test were being passed through. This
4699 no longer happens. The underlying message is available in $acl_verify_
4700 message for both "message" and "log_message" expansions, so it can be
4701 passed through if needed.
4703 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4704 new expansion $bh_ to give the encoded byte string without charset
4705 translation. Translation happens only if iconv() is available; HAVE_ICONV
4706 indicates this at build time. HEADERS_CHARSET gives the charset to
4707 translate to; headers_charset can change it in the configuration, and
4708 "headers charset" can change it in an individual filter file.
4710 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4711 that creates RFC 2047 encoded "words" labels them as that charset instead
4712 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4713 expansion operator; (ii) when Exim creates a From: line for a local
4714 message; (iii) when a header line is rewritten to include a "phrase" part.
4716 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4717 buggy, causing it to skip the first lines of messages whose message ID
4718 ended in 'D'. This would not have bitten before Exim release 4.14, because
4719 message IDs were unlikely to end in 'D' before then. The effect was to have
4720 incorrect size information for certain domains.
4722 45. #include "config.h" was missing at the start of the crypt16.c module. This
4723 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4726 46. If there was a timeout during a "random" callout check, Exim treated it as
4727 a failure of the random address, and carried on sending RSET and the real
4728 address. If the delay was just some slowness somewhere, the response to the
4729 original RCPT would be taken as a response to RSET and so on, causing
4730 mayhem of various kinds.
4732 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4733 when I implemented it. It didn't allow for the fact that some option values
4734 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4735 the right test for positive values.
4737 48. Domain names in DNS records are case-independent. Exim always looks them up
4738 in lower case. Some resolvers return domain names in exactly the case they
4739 appear in the zone file, that is, they may contain uppercase letters. Not
4740 all resolvers do this - some return always lower case. Exim was treating a
4741 change of case by a resolver as a change of domain, similar to a widening
4742 of a domain abbreviation. This triggered its re-routing code and so it was
4743 trying to route what was effectively the same domain again. This normally
4744 caused routing to fail (because the router wouldn't handle the domain
4745 twice). Now Exim checks for this case specially, and just changes the
4746 casing of the domain that it ultimately uses when it transmits the message
4749 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4752 50. If a filter generated a file delivery with a non-absolute name (possible if
4753 no home directory exists for the router), the forbid_file option was not
4756 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4757 the existing equality matching.
4759 52. Exim was using ints instead of ino_t variables in some places where it was
4760 dealing with inode numbers.
4762 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4763 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4764 called TMPDIR, and if it finds it is different, it changes its value.
4766 54. The smtp_printf() function is now made available to local_scan() so
4767 additional output lines can be written before returning. There is also an
4768 smtp_fflush() function to enable the detection of a dropped connection.
4769 The variables smtp_input and smtp_batched_input are exported to
4772 55. Changed the default runtime configuration: the message "Unknown user"
4773 has been removed from the ACL, and instead placed on the localuser router,
4774 using the cannot_route_message feature. This means that any verification
4775 failures that generate their own messages won't get overridden. Similarly,
4776 the "Unrouteable address" message that was in the ACL for unverifiable
4777 relay addresses has also been removed.
4779 56. Added hosts_avoid_esmtp to the smtp transport.
4781 57. The exicyclog script was not checking for the esoteric option
4782 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4783 will work only if exicyclog is run under the appropriate euid.
4785 58. Following a discussion on the list, the rules by which Exim recognises line
4786 endings on incoming messages have been changed. The -dropcr and drop_cr
4787 options are now no-ops, retained only for backwards compatibility. The
4788 following line terminators are recognized: LF CRLF CR. However, special
4789 processing applies to CR:
4791 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4792 nor a local message in the state where . is a terminator.
4794 (ii) If a bare CR is encountered in a header line, an extra space is added
4795 after the line terminator so as not to end the header. The reasoning
4796 behind this is that bare CRs in header lines are most likely either
4797 to be mistakes, or people trying to play silly games.
4799 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4800 was being incorrectly given as 18 bytes larger than it should have been.
4801 This is a VOB (very old bug).
4803 60. This may never have affected anything current, but just in case it has:
4804 When the local host is found other than at the start of a list of hosts,
4805 the local host, those with the same MX, and any that follow, are discarded.
4806 When the list in question was part of a longer list of hosts, the following
4807 hosts (not currently being processed) were also being discarded. This no
4808 longer happens. I'm not sure if this situation could ever has previously
4811 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4814 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4815 header. This is something that is recommended in a new Internet Draft, and
4816 is something that is documented as being done by Sendmail. There are two
4817 possible values. For messages generated by the autoreply transport, Exim
4820 Auto-Submitted: auto-replied
4822 whereas for all other generated messages (e.g. bounces) it adds
4824 Auto-Submitted: auto-generated
4826 63. The "personal" condition in filters now includes a test for the
4827 Auto-Submitted: header. If it contains the string "auto-" the message it
4828 not considered personal.
4830 64. Added rcpt_include_affixes as a generic transport option.
4832 65. Added queue_only_override (default true).
4834 66. Added the syslog_duplication option.
4836 67. If what should have been the first header line of a message consisted of
4837 a space followed by a colon, Exim was mis-interpreting it as a header line.
4838 It isn't of course - it is syntactically invalid and should therefore be
4839 treated as the start of the message body. The misbehaviour could have
4840 caused a number of strange effects, including loss of data in subsequent
4841 header lines, and spool format errors.
4843 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4844 client host had authenticated. This control can now be exercised by an ACL
4845 for more flexibility.
4847 69. By default, callouts do not happen when testing with -bh. There is now a
4848 variant, -bhc, which does actually run the callout code, including
4849 consulting and updating the callout cache.
4851 70. Added support for saslauthd authentication, courtesy of Alexander
4854 71. If statvfs() failed on the spool or log directories while checking their
4855 size for availability, Exim confusingly gave the error "space shortage".
4856 Furthermore, in debugging mode it crashed with a floating point exception.
4857 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4858 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4859 serious problem, Exim now writes to the main and panic logs when this
4860 happens, with details of the failure. It then refuses to accept the
4861 incoming message, giving the message "spool directory problem" or "log
4862 directory problem" with a 421 code for SMTP messages.
4864 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4865 0, 1, and 2 exist, because some OS complain for execs without them (see
4866 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4867 descriptors. However, the code omitted to check that the open succeeded,
4868 causing mysterious errors if for some reason the permissions on /dev/null
4869 got screwed. Now Exim writes a message to the main and panic logs, and
4870 bombs out if it can't open /dev/null.
4872 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4873 interact so that it is all more flexible. It is supposed to remain
4874 backwards compatible. Also added extra_local_interfaces.
4876 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4877 to bomb out with an assertion failure - to the client this appears as a
4878 connection drop. This problem occurs in the part of the code that was taken
4879 from the Samba project. Fortunately, the assertion is in a very simple
4880 function, so I have fixed this by reproducing the function inline in the
4881 one place where it is called, and arranging for authentication to fail
4882 instead of killing the process with assert().
4884 75. The SPA client code was not working when the server requested OEM rather
4885 than Unicode encoding.
4887 76. Added code to make require_files with a specific uid setting more usable in
4888 the case where statting the file as root fails - usually a non-root-mounted
4889 NFS file system. When this happens and the failure is EACCES, Exim now
4890 forks a subprocess and does the per-uid checking as the relevant uid.
4892 77. Added process_log_path.
4894 78. If log_file_path was not explicitly set, a setting of check_log_space or
4895 check_log_inodes was ignored.
4897 79. If a space check for the spool or log partitions fails, the incident is now
4898 logged. Of course, in the latter case the data may get lost...
4900 80. Added the %p formatting code to string_format() so that it can be used to
4901 print addresses in debug_print(). Adjusted all the address printing in the
4902 debugging in store.c to use %p rather than %d.
4904 81. There was a concern that a line of code in smtp_in.c could overflow a
4905 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4906 initially expressed, the concern was not well-founded, because trailing
4907 spaces are removed early. However, if the trailing spaces were followed by
4908 a NULL, they did not get removed, so the overflow was possible. Two fixes
4911 (a) I re-wrote the offending code in a cleaner fashion.
4912 (b) If an incoming SMTP command contains a NULL character, it is rejected
4915 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4916 runs initgroups(), so that if the Exim user is in any additional groups,
4917 they will be used during message reception.
4923 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob