1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.552 2008/09/03 18:53:29 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.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
72 TK/01 Add preliminary DKIM support. Currently requires a forked version of
73 ALT-N's libdkim that I have put here:
74 http://duncanthrax.net/exim-experimental/
76 Note to Michael Haardt: I had to rename some vars in sieve.c. They
77 were called 'true' and it seems that C99 defines that as a reserved
78 keyword to be used with 'bool' variable types. That means you could
79 not include C99-style headers which use bools without triggering
80 build errors in sieve.c.
82 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
83 as mailq or other aliases. Changed the --help handling significantly
84 to do whats expected. exim_usage() emits usage/help information.
86 SC/01 Added the -bylocaldomain option to eximstats.
88 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
90 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
92 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
98 PH/01 Another patch from the Sieve maintainer.
100 PH/02 When an IPv6 address is converted to a string for single-key lookup
101 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
102 dots are used instead of colons so that keys in lsearch files need not
103 contain colons. This was done some time before quoting was made available
104 in lsearch files. However, iplsearch files do require colons in IPv6 keys
105 (notated using the quote facility) so as to distinguish them from IPv4
106 keys. This meant that lookups for IP addresses in host lists did not work
107 for iplsearch lookups.
109 This has been fixed by arranging for IPv6 addresses to be expressed with
110 colons if the lookup type is iplsearch. This is not incompatible, because
111 previously such lookups could never work.
113 The situation is now rather anomolous, since one *can* have colons in
114 ordinary lsearch keys. However, making the change in all cases is
115 incompatible and would probably break a number of configurations.
117 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
120 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
121 conversion specification without a maximum field width, thereby enabling
122 a rogue spamd server to cause a buffer overflow. While nobody in their
123 right mind would setup Exim to query an untrusted spamd server, an
124 attacker that gains access to a server running spamd could potentially
125 exploit this vulnerability to run arbitrary code as the Exim user.
127 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
128 $primary_hostname instead of what libspf2 thinks the hosts name is.
130 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
131 a directory entry by the name of the lookup key. Previously, if a
132 symlink pointed to a non-existing file or a file in a directory that
133 Exim lacked permissions to read, a lookup for a key matching that
134 symlink would fail. Now it is enough that a matching directory entry
135 exists, symlink or not. (Bugzilla 503.)
137 PH/03 The body_linecount and body_zerocount variables are now exported in the
140 PH/04 Added the $dnslist_matched variable.
142 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
143 This means they are set thereafter only if the connection becomes
146 PH/06 Added the client_condition to authenticators so that some can be skipped
147 by clients under certain conditions.
149 PH/07 The error message for a badly-placed control=no_multiline_responses left
150 "_responses" off the end of the name.
152 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
154 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
155 (without spaces) instead of just copying the configuration text.
157 PH/10 Added the /noupdate option to the ratelimit ACL condition.
159 PH/11 Added $max_received_linelength.
161 PH/12 Added +ignore_defer and +include_defer to host lists.
163 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
164 way in which PCRE > 7.0 is built.
166 PH/14 Implemented queue_only_load_latch.
168 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
169 MAIL command. The effect was to mangle the value on 64-bit systems.
171 PH/16 Another patch from the Sieve maintainer.
173 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
175 PH/18 If a system quota error occurred while trying to create the file for
176 a maildir delivery, the message "Mailbox is full" was not appended to the
177 bounce if the delivery eventually timed out. Change 4.67/27 below applied
178 only to a quota excession during the actual writing of the file.
180 PH/19 It seems that peer DN values may contain newlines (and other non-printing
181 characters?) which causes problems in log lines. The DN values are now
182 passed through string_printing() before being added to log lines.
184 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
185 and InterBase are left for another time.)
187 PH/21 Added message_body_newlines option.
189 PH/22 Guard against possible overflow in moan_check_errorcopy().
191 PH/23 POSIX allows open() to be a macro; guard against that.
193 PH/24 If the recipient of an error message contained an @ in the local part
194 (suitably quoted, of course), incorrect values were put in $domain and
195 $local_part during the evaluation of errors_copy.
201 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
202 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
205 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
206 issue a MAIL command.
208 PH/02 In an ACL statement such as
210 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
212 if a client was not listed at all, or was listed with a value other than
213 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
214 the condition was not true (as it should be), so access was not denied.
215 The bug was that the ! inversion was incorrectly passed on to the second
216 item. This has been fixed.
218 PH/03 Added additional dnslists conditions == and =& which are different from
219 = and & when the dns lookup returns more than one IP address.
221 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
222 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
224 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
225 FSYNC, which compiles an option called disable_fsync that allows for
226 bypassing fsync(). The documentation is heavily laced with warnings.
228 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
230 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
231 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
232 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
233 including adding "make clean"; (3) Added -fPIC when compiling the test
234 dynamically loaded module, to get rid of a warning.
236 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
237 message fails, move_frozen_messages = true and ignore_bounce_errors_after
238 = 0s. The bug is otherwise harmless.
240 PH/07 There was a bug in the dovecot authenticator such that the value of
241 $auth1 could be overwritten, and so not correctly preserved, after a
242 successful authentication. This usually meant that the value preserved by
243 the server_setid option was incorrect.
245 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
247 PH/09 Installed PCRE release 7.0.
249 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
250 run for batched SMTP input. It is now run at the start of every message
251 in the batch. While fixing this I discovered that the process information
252 (output by running exiwhat) was not always getting set for -bs and -bS
253 input. This is fixed, and it now also says "batched" for BSMTP.
255 PH/11 Added control=no_pipelining.
257 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
258 patch, slightly modified), and move the expansion of helo_data till after
259 the connection is made in the smtp transport (so it can use these
262 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
264 PH/14 Added log_selector = +pid.
266 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
268 PH/16 Add ${if forany and ${if forall.
270 PH/17 Added dsn_from option to vary the From: line in DSNs.
272 PH/18 Flush SMTP output before performing a callout, unless control =
273 no_callout_flush is set.
275 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
276 was true (the default) a successful delivery failed to delete the retry
277 item, thus causing premature timeout of the address. The bug is now
280 PH/20 Added hosts_avoid_pipelining to the smtp transport.
282 PH/21 Long custom messages for fakedefer and fakereject are now split up
283 into multiline reponses in the same way that messages for "deny" and
284 other ACL rejections are.
286 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
287 with slight modification.
289 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
290 draft, changing the syntax and factoring some duplicate code".
292 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
293 for deliveries of the second and subsequent messages over the same SMTP
296 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
297 ${reduce, with only minor "tidies".
299 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
301 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
302 expansion side effects.
304 PH/27 When a message times out after an over-quota error from an Exim-imposed
305 quota, the bounce message says "mailbox is full". This message was not
306 being given when it was a system quota that was exceeded. It now should
309 MH/03 Made $recipients available in local_scan(). local_scan() already has
310 better access to the recipient list through recipients_list[], but
311 $recipients can be useful in postmaster-provided expansion strings.
313 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
314 in the case of a MAIL command with additional options following the
315 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
316 were accidentally chopped off.
318 PH/29 SMTP synchronization checks are implemented when a command is read -
319 there is a check that no more input is waiting when there shouldn't be
320 any. However, for some commands, a delay in an ACL can mean that it is
321 some time before the response is written. In this time, more input might
322 arrive, invalidly. So now there are extra checks after an ACL has run for
323 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
324 pipelining has not been advertised.
326 PH/30 MH's patch to allow iscntrl() characters to be list separators.
328 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
329 returned in the SMTP response when smtp_return_error_details was false.
332 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
333 instead of the C I/O that was originally supplied, because problems were
336 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
337 Exim which did not show up earlier: it was assuming that a call to
338 SSL_CTX_set_info_callback() might give an error value. In fact, there is
339 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
340 was a macro that became an assignment, so it seemed to work. This has
341 changed to a proper function call with a void return, hence the compile
342 error. Exim's code has been fixed.
344 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
347 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
349 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
351 PH/37 If a message is not accepted after it has had an id assigned (e.g.
352 because it turns out to be too big or there is a timeout) there is no
353 "Completed" line in the log. When some messages of this type were
354 selected by exigrep, they were listed as "not completed". Others were
355 picked up by some special patterns. I have improved the selection
356 criteria to be more general.
358 PH/38 The host_find_failed option in the manualroute router can now be set
359 to "ignore", to completely ignore a host whose IP address cannot be
360 found. If all hosts are ignored, the behaviour is controlled by the new
361 host_all_ignored option.
363 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
364 homing or because of multiple MX records with /mx) generated more than
365 one IP address, and the following item turned out to be the local host,
366 all the secondary addresses of the first item were incorrectly removed
367 from the list, along with the local host and any following hosts (which
368 is what is supposed to happen).
370 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
371 whoever called Exim into the -H file. In the case of the daemon it was
372 behaving confusingly. When first started, it used values for whoever
373 started the daemon, but after a SIGHUP it used the Exim user (because it
374 calls itself on a restart). I have changed the code so that it now always
377 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
378 message are rejected with the same error (e.g. no authentication or bad
379 sender address), and a DATA command is nevertheless sent (as can happen
380 with PIPELINING or a stupid MUA), the error message that was given to the
381 RCPT commands is included in the rejection of the DATA command. This is
382 intended to be helpful for MUAs that show only the final error to their
385 PH/42 Another patch from the Sieve maintainer.
387 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
388 Eximstats - Fixed some broken HTML links and added missing column headers
390 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
391 columns for Rejects, Temp Rejects, Ham, and Spam rows.
393 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
395 PH/43 Yet another patch from the Sieve maintainer.
397 PH/44 I found a way to check for a TCP/IP connection going away before sending
398 the response to the final '.' that terminates a message, but only in the
399 case where the client has not sent further data following the '.'
400 (unfortunately, this is allowed). However, in many cases there won't be
401 any further data because there won't be any more messages to send. A call
402 to select() can be used: if it shows that the input is "ready", there is
403 either input waiting, or the socket has been closed. An attempt to read
404 the next input character can distinguish the two cases. Previously, Exim
405 would have sent an OK response which the client would never have see.
406 This could lead to message repetition. This fix should cure that, at
407 least in a lot of common cases.
409 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
410 advertised in response to EHLO.
416 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
417 fixed by 4.65/MH/01 (is this a record?) are fixed:
419 (i) An empty string was always treated as zero by the numeric comparison
420 operators. This behaviour has been restored.
422 (ii) It is documented that the numeric comparison operators always treat
423 their arguments as decimal numbers. This was broken in that numbers
424 starting with 0 were being interpreted as octal.
426 While fixing these problems I realized that there was another issue that
427 hadn't been noticed. Values of message_size_limit (both the global option
428 and the transport option) were treated as octal if they started with 0.
429 The documentation was vague. These values are now always treated as
430 decimal, and I will make that clear in the documentation.
436 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
437 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
440 MH/01 Don't check that the operands of numeric comparison operators are
441 integers when their expansion is in "skipping" mode (fixes bug
442 introduced by 4.64-PH/07).
444 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
445 child addresses, Exim now panics and dies. Previously, because the count
446 is held in a short int, deliveries were likely to be lost. As such a
447 large number of recipients for a single message is ridiculous
448 (performance will be very, very poor), I have chosen to impose a limit
449 rather than extend the field.
455 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
456 leftover -K file (the existence of which was triggered by #402).
457 While we were at it, introduced process PID as part of the -K
458 filename. This should rule out race conditions when creating
461 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
462 processing considerably. Previous code took too long for large mails,
463 triggering a timeout which in turn triggers #401.
465 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
466 in the DK code in transports.c. sendfile() is not really portable,
467 hence the _LINUX specificness.
469 TF/01 In the add_headers option to the mail command in an Exim filter,
470 there was a bug that Exim would claim a syntax error in any
471 header after the first one which had an odd number of characters
474 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
475 callout verification, Exim cached a "reject" for the entire domain. This
476 is correct for most verifications, but it is not correct for a recipient
477 verification with use_sender or use_postmaster set, because in that case
478 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
479 case of MAIL FROM:<> rejection from other early rejections (e.g.
480 rejection of HELO). When verifying a recipient using a non-null MAIL
481 address, the cache is ignored if it shows MAIL FROM:<> rejection.
482 Whatever the result of the callout, the value of the domain cache is
483 left unchanged (for any other kind of callout, getting as far as trying
484 RCPT means that the domain itself is ok).
486 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
489 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
490 manifest itself as EPIPE rather than ECONNECT. When tidying away a
491 session, the daemon ignores ECONNECT errors and logs others; it now
492 ignores EPIPE as well.
494 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
495 (quoted-printable decoding).
497 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
498 later the small subsequent patch to fix an introduced bug.
500 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
502 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
504 PH/08 An error is now given if message_size_limit is specified negative.
506 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
507 to be given (somewhat) arbitrary names.
509 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
512 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
513 miscellaneous code fixes
515 PH/10 Added the log_reject_target ACL modifier to specify where to log
518 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
519 hostname. This is wrong, because it relates to the incoming message (and
520 probably the interface on which it is arriving) and not to the outgoing
521 callout (which could be using a different interface). This has been
522 changed to use the value of the helo_data option from the smtp transport
523 instead - this is what is used when a message is actually being sent. If
524 there is no remote transport (possible with a router that sets up host
525 addresses), $smtp_active_hostname is used.
527 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
528 tweaks were necessary in order to get it to work (see also 21 below):
529 (a) The code assumed that strncpy() returns a negative number on buffer
530 overflow, which isn't the case. Replaced with Exim's string_format()
532 (b) There were several signed/unsigned issues. I just did the minimum
533 hacking in of casts. There is scope for a larger refactoring.
534 (c) The code used strcasecmp() which is not a standard C function.
535 Replaced with Exim's strcmpic() function.
536 (d) The code set only $1; it now sets $auth1 as well.
537 (e) A simple test gave the error "authentication client didn't specify
538 service in request". It would seem that Dovecot has changed its
539 interface. Fortunately there's a specification; I followed it and
540 changed what the client sends and it appears to be working now.
542 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
545 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
546 address A is aliased to B and C, where B exists and C does not. Without
547 -v the output is "A verified" because verification stops after a
548 successful redirection if more than one address is generated. However,
549 with -v the child addresses are also verified. Exim was outputting "A
550 failed to verify" and then showing the successful verification for C,
551 with its parentage. It now outputs "B failed to verify", showing B's
552 parentage before showing the successful verification of C.
554 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
555 look up a TXT record in a specific list after matching in a combined
558 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
559 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
560 they consult the DNS. I had assumed they would set it the way they
561 wanted; and indeed my experiments on Linux seem to show that in some
562 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
563 To be on the safe side, however, I have now made the interface to
564 host_find_byname() similar to host_find_bydns(), with an argument
565 containing the DNS resolver options. The host_find_byname() function now
566 sets these options at its start, just as host_find_bydns() does. The smtp
567 transport options dns_qualify_single and dns_search_parents are passed to
568 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
569 of host_find_byname() use the default settings of RES_DEFNAMES
570 (qualify_single) but not RES_DNSRCH (search_parents).
572 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
573 spool_read_header() do less string testing, by means of a preliminary
574 switch on the second character of optional "-foo" lines. (This is
575 overdue, caused by the large number of possibilities that now exist.
576 Originally there were few.) While I was there, I also converted the
577 str(n)cmp tests so they don't re-test the leading "-" and the first
578 character, in the hope this might squeeze out yet more improvement.
580 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
581 flag allowing group syntax was set by the header_syntax check but not
582 turned off, possible causing trouble later; (2) The flag was not being
583 set at all for the header_verify test, causing "group"-style headers to
584 be rejected. I have now set it in this case, and also caused header_
585 verify to ignore an empty address taken from a group. While doing this, I
586 came across some other cases where the code for allowing group syntax
587 while scanning a header line wasn't quite right (mostly, not resetting
588 the flag correctly in the right place). These bugs could have caused
589 trouble for malformed header lines. I hope it is now all correct.
591 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
592 with the "reply" argument non-NULL. The code, however (which originally
593 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
594 but it didn't always do it. This confused somebody who was copying the
595 code for some other use. I have removed all the tests.
597 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
598 feature that was used to support insecure browsers during the U.S. crypto
599 embargo. It requires special client support, and Exim is probably the
600 only MTA that supported it -- and would never use it because real RSA is
601 always available. This code has been removed, because it had the bad
602 effect of slowing Exim down by computing (never used) parameters for the
603 RSA_EXPORT functionality.
605 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
606 authenticator to fail if there's a tab character in the incoming data
607 (there should never be unless someone is messing about, as it's supposed
608 to be base64-encoded). Also added, on Timo's advice, the "secured" option
609 if the connection is using TLS or if the remote IP is the same as the
610 local IP, and the "valid-client-cert option" if a client certificate has
613 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
614 authenticators. This can be used for authorization after authentication
615 succeeds. (In the case of plaintext, it servers for both authentication
618 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
619 if any retry times were supplied.
621 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
622 connection, where there is no client IP address to check. In this
623 situation, the verify now always succeeds.
625 PH/25 Applied John Jetmore's -Mset patch.
627 PH/26 Added -bem to be like -Mset, but loading a message from a file.
629 PH/27 In a string expansion for a processed (not raw) header when multiple
630 headers of the same name were present, leading whitespace was being
631 removed from all of them, but trailing whitespace was being removed only
632 from the last one. Now trailing whitespace is removed from each header
633 before concatenation. Completely empty headers in a concatenation (as
636 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
637 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
639 PH/29 [Removed. This was a change that I later backed out, and forgot to
640 correct the ChangeLog entry (that I had efficiently created) before
641 committing the later change.]
643 PH/30 Exim was sometimes attempting to deliver messages that had suffered
644 address errors (4xx response to RCPT) over the same connection as other
645 messages routed to the same hosts. Such deliveries are always "forced",
646 so retry times are not inspected. This resulted in far too many retries
647 for the affected addresses. The effect occurred only when there were more
648 hosts than the hosts_max_try setting in the smtp transport when it had
649 the 4xx errors. Those hosts that it had tried were not added to the list
650 of hosts for which the message was waiting, so if all were tried, there
651 was no problem. Two fixes have been applied:
653 (i) If there are any address or message errors in an SMTP delivery, none
654 of the hosts (tried or untried) are now added to the list of hosts
655 for which the message is waiting, so the message should not be a
656 candidate for sending over the same connection that was used for a
657 successful delivery of some other message. This seems entirely
658 reasonable: after all the message is NOT "waiting for some host".
659 This is so "obvious" that I'm not sure why it wasn't done
660 previously. Hope I haven't missed anything, but it can't do any
661 harm, as the worst effect is to miss an optimization.
663 (ii) If, despite (i), such a delivery is accidentally attempted, the
664 routing retry time is respected, so at least it doesn't keep
665 hammering the server.
667 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
668 in ${readsocket because some servers need this prod.
670 PH/32 Added some extra debug output when updating a wait-xxx database.
672 PH/33 The hint "could be header name not terminated by colon", which has been
673 given for certain expansion errors for a long time, was not being given
674 for the ${if def:h_colon_omitted{... case.
676 PH/34 The spec says: "With one important exception, whenever a domain list is
677 being scanned, $domain contains the subject domain." There was at least
678 one case where this was not true.
680 PH/35 The error "getsockname() failed: connection reset by peer" was being
681 written to the panic log as well as the main log, but it isn't really
682 panic-worthy as it just means the connection died rather early on. I have
683 removed the panic log writing for the ECONNRESET error when getsockname()
686 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
687 runs only) independently of the message's sender address. This meant
688 that, if the 4xx error was in fact related to the sender, a different
689 message to the same recipient with a different sender could confuse
690 things. In particualar, this can happen when sending to a greylisting
691 server, but other circumstances could also provoke similar problems.
692 I have changed the default so that the retry time for these errors is now
693 based a combination of the sender and recipient addresses. This change
694 can be overridden by setting address_retry_include_sender=false in the
697 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
698 remote server are returned as part of bounce messages. This was not
699 happening for LMTP over a pipe (the lmtp transport), but now it is the
700 same for both kinds of LMTP.
702 PH/38 Despite being documented as not happening, Exim was rewriting addresses
703 in header lines that were in fact CNAMEs. This is no longer the case.
705 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
706 and queue runs started by the daemon processed all messages. This has
707 been fixed so that -R and -S can now usefully be given with -q<time>.
709 PH/40 Import PCRE release 6.7 (fixes some bugs).
711 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
713 PH/42 Give an error if -q is specified more than once.
715 PH/43 Renamed the variables $interface_address and $interface_port as
716 $received_ip_address and $received_port, to make it clear that these
717 values apply to message reception, and not to the outgoing interface when
718 a message is delivered. (The old names remain recognized, of course.)
720 PH/44 There was no timeout on the connect() call when using a Unix domain
721 socket in the ${readsocket expansion. There now is.
723 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
724 be meaningful with "accept".
726 SC/01 Eximstats V1.43
727 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
729 SC/02 Eximstats V1.44
730 Use a glob alias rather than an array ref in the generated
731 parser. This improves both readability and performance.
733 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
734 Collect SpamAssassin and rejection statistics.
735 Don't display local sender or destination tables unless
736 there is data to show.
737 Added average volumes into the top table text output.
739 SC/04 Eximstats V1.46
740 Collect data on the number of addresses (recipients)
741 as well as the number of messages.
743 SC/05 Eximstats V1.47
744 Added 'Message too big' to the list of mail rejection
745 reasons (thanks to Marco Gaiarin).
747 SC/06 Eximstats V1.48
748 Mainlog lines which have GMT offsets and are too short to
749 have a flag are now skipped.
751 SC/07 Eximstats V1.49 (Alain Williams)
752 Added the -emptyok flag.
754 SC/08 Eximstats V1.50
755 Fixes for obtaining the IP address from reject messages.
757 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
758 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
759 whitesspace changes from 4.64-PH/27
761 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
764 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
765 are found, allow negative numbers in numeric criteria)
767 JJ/06 exipick.20061117.2, added new $message_body_missing variable
769 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
770 to match changes made in 4.64-PH/43
772 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
774 PH/47 Put in an explicit test for a DNS lookup of an address record where the
775 "domain" is actually an IP address, and force a failure. This locks out
776 those revolvers/nameservers that support "A-for-A" lookups, in
777 contravention of the specifications.
779 PH/48 When a host name was looked up from an IP address, and the subsequent
780 forward lookup of the name timed out, the host name was left in
781 $sender_host_name, contrary to the specification.
783 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
784 restricted to single-key lookups, Exim was not diagnosing an error if
785 * or *@ was used with a query-style lookup.
787 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
789 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
790 long ago, but noone interested enough thought of it. Let's just say that
791 the "1.1" means that there are some new functions that weren't there at
792 some point in the past.
794 PH/51 Error processing for expansion failure of helo_data from an smtp
795 transport during callout processing was broken.
797 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
798 tested/used via the -bh/-bhc/-bs options.
800 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
801 bug, fixed in subsequent PCRE releases).
803 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
804 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
806 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
812 SC/01 Use a glob alias rather than an array ref in eximstats generated
813 parser. This improves both readability and performance.
815 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
816 Don't display local sender or destination tables in eximstats unless
817 there is data to show.
818 Added average volumes into the eximstats top table text output.
820 SC/03 Collect data on the number of addresses (recipients) as well
821 as the number of messages in eximstats.
823 TF/01 Correct an error in the documentation for the redirect router. Exim
824 does (usually) call initgroups() when daemonizing.
826 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
827 with consistent privilege compared to when running as a daemon.
829 TF/03 Note in the spec that $authenticated_id is not set for local
830 submissions from trusted users.
832 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
833 Thanks to Dean Brooks <dean@iglou.com> for the patch.
835 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
836 by adding some example configuration directives to the default
837 configuration file. A little bit of work is required to uncomment the
838 directives and define how usernames and passwords are checked, but
839 there is now a framework to start from.
841 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
842 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
843 without this. I don't know how relevant this is to other LDAP libraries.
845 PH/02 Add the verb name to the "unknown ACL verb" error.
847 PH/03 Magnus Holmgren's patch for filter_prepend_home.
849 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
851 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
852 directory not expanded when it should be if an expanded home directory
853 was set for the address (which is overridden by the transport).
855 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
858 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
859 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
860 because it is too late at that time, and has no effect.
862 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
863 security issue with \' (bugzilla #107). I could not use the
864 PQescapeStringConn() function, because it needs a PGconn value as one of
867 PH/08 When testing addresses using -bt, indicate those final addresses that
868 are duplicates that would not cause an additional delivery. At least one
869 person was confused, thinking that -bt output corresponded to deliveries.
870 (Suppressing duplicates isn't a good idea as you lose the information
871 about possibly different redirections that led to the duplicates.)
873 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
874 systems where poll() doesn't work, in particular OS X.
876 PH/10 Added more information to debugging output for retry time not reached.
878 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
879 operations in malware.c.
881 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
884 PH/13 If write_rejectlog was set false when logging was sent to syslog with
885 syslog_duplication set false, log lines that would normally be written
886 both the the main log and to the reject log were not written to syslog at
889 PH/14 In the default configuration, change the use of "message" in ACL warn
890 statements to "add_header".
892 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
893 not followed by a command (e.g. "seen endif").
895 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
896 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
899 PH/17 Added extra conditions to the default value of delay_warning_condition
903 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
904 { match{$h_precedence:}{(?i)bulk|list|junk} } \
905 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
908 The Auto-Submitted: and various List- headers are standardised, whereas I
909 don't think Precedence: ever was.
911 PH/18 Refactored debugging code in route_finduser() to show more information,
912 in particular, the error code if getpwnam() issues one.
914 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
915 This is apparently needed in addition to the PH/07 change above to avoid
916 any possible encoding problems.
918 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
919 but not after initializing Perl.
921 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
922 output them only if debugging. By default they are written stderr,
923 apparently, which is not desirable.
925 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
928 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
931 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
933 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
934 authenticated or an ident call has been made. Suppress the default
935 values for $authenticated_id and $authenticated_sender (but permit -oMai
936 and -oMas) when testing with -bh.
938 PH/24 Re-jigged the order of the tests in the default configuration so that the
939 tests for valid domains and recipients precede the DNS black list and CSA
940 tests, on the grounds that those ones are more expensive.
942 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
943 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
944 This bug exists in every version of Exim that I still have, right back to
947 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
948 However, an attempt to turn on case-sensitivity in a regex key by
949 including (?-i) didn't work because the subject string was already
950 lowercased, and the effects were non-intuitive. It turns out that a
951 one-line patch can be used to allow (?-i) to work as expected.
957 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
958 other effects) broke the use of negated acl sub-conditions.
960 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
963 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
964 "Deny" causes Exim to reject the incoming connection with a 554 error.
965 Unfortunately, if there is a major crisis, such as a disk failure,
966 tcp-wrappers gives "deny", whereas what one would like would be some
967 kind of temporary error. A kludge has been added to help with this.
968 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
969 554 error is used if errno is still zero or contains ENOENT (which occurs
970 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
973 PH/03 Add -lutil to the default FreeBSD LIBS setting.
975 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
976 errors. Otherwise a message that provokes a temporary error (when other
977 messages do not) can cause a whole host to time out.
979 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
980 addresses were routed directly to files or pipes from a redirect router.
981 File deliveries just didn't batch; pipe deliveries might have suffered
984 PH/06 A failure to get a lock for a hints database would erroneously always say
985 "Failed to get write lock", even when it was really a read lock.
987 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
988 of 0600. This has been changed to use the value of the lockfile_mode
989 option (which defaults to 0600).
991 PH/08 Applied small patch from the Sieve maintainer.
993 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
994 folder from quota calculations, a direct delivery into this folder messed
995 up the contents of the maildirsize file. This was because the regex was
996 used only to exclude .Trash (or whatever) when the size of the mailbox
997 was calculated. There was no check that a delivery was happening into an
998 excluded directory. This bug has been fixed by ignoring all quota
999 processing for deliveries into excluded directories.
1001 PH/10 Added the maildirfolder_create_regex option to appendfile.
1007 PH/01 The code for finding all the local interface addresses on a FreeBSD
1008 system running IPv6 was broken. This may well have applied to all BSD
1009 systems, as well as to others that have similar system calls. The broken
1010 code found IPv4 interfaces correctly, but gave incorrect values for the
1011 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1012 that it would not match correctly against @[] and not recognize the IPv6
1015 PH/02 The ipliteral router was not recognizing addresses of the form user@
1016 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1018 PH/03 Added disable_ipv6.
1020 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1021 decoded form, because it is most often used to construct To: headers
1022 lines in autoreplies, and the decoded form may well be syntactically
1023 invalid. However, $reply_address has leading white space removed, and all
1024 newlines turned into spaces so that the autoreply transport does not
1027 PH/05 If group was specified without a user on a router, and no group or user
1028 was specified on a transport, the group from the router was ignored.
1030 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1031 for visible compile-time settings that can be used to change these
1032 numbers, for those that want even more. Backwards compatibility with old
1033 spool files has been maintained. However, going back to a previous Exim
1034 release will lost any variables that are in spool files.
1036 PH/07 Two small changes when running in the test harness: increase delay when
1037 passing a TCP/IP connection to a new process, in case the original
1038 process has to generate a bounce, and remove special handling of
1039 127.0.0.2 (sic), which is no longer necessary.
1041 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1042 be the same on different OS.
1044 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1047 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1048 whether --show-vars was specified or not
1050 JJ/02 exipick: Added support for new ACL variable spool format introduced
1053 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1054 syntactically invalid From: or Reply-to: line, and a filter used this to
1055 generate an autoreply, and therefore failed to obtain an address for the
1056 autoreply, Exim could try to deliver to a non-existent relative file
1057 name, causing unrelated and misleading errors. What now happens is that
1058 it logs this as a hard delivery error, but does not attempt to create a
1061 PH/11 The exinext utility has a -C option for testing purposes, but although
1062 the given file was scanned by exinext itself; it wasn't being passed on
1063 when Exim was called.
1065 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1066 an end-of-file indication when reading a command response.
1068 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1069 compiled. In many other places in Exim, IPv6 addresses are always
1070 recognized, so I have changed this. It also means that IPv4 domain
1071 literals of the form [IPV4:n.n.n.n] are now always recognized.
1073 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1074 used if the router is not running as root, for example, when verifying at
1075 ACL time, or when using -bh. The debugging output from this situation was
1076 non-existent - all you got was a failure to exec. I have made two
1079 (a) Failures to set uid/gid, the current directory, or a process leader
1080 in a subprocess such as that created by queryprogram now generate
1081 suitable debugging ouput when -d is set.
1083 (b) The queryprogram router detects when it is not running as root,
1084 outputs suitable debugging information if -d is set, and then runs
1085 the subprocess without attempting to change uid/gid.
1087 PH/15 Minor change to Makefile for building test_host (undocumented testing
1090 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1091 additional section of a DNS packet that returns MX or SRV records.
1092 Instead, it always explicitly searches for A/AAAA records. This avoids
1093 major problems that occur when a DNS server includes only records of one
1094 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1095 fixed another bug: if SRV records were looked up and the corresponding
1096 address records were *not* found in the additional section, the port
1097 values from the SRV records were lost.
1099 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1100 using the correct key (the original address) when searching the retry
1101 rules in order to find which one to use for generating the retry hint.
1103 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1104 adding the default one. Similarly, if it contains a Reply-To: header, the
1105 errors_reply_to option, if set, is not used.
1107 PH/19 When calculating a retry time, Exim used to measure the "time since
1108 failure" by looking at the "first failed" field in the retry record. Now
1109 it does not use this if it is later than than the arrival time of the
1110 message. Instead it uses the arrival time. This makes for better
1111 behaviour in cases where some deliveries succeed, thus re-setting the
1112 "first failed" field. An example is a quota failure for a huge message
1113 when small messages continue to be delivered. Without this change, the
1114 "time since failure" will always be short, possible causing more frequent
1115 delivery attempts for the huge message than are intended.
1116 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1118 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1119 $1, $2, $3) because the numerical variables can be reset during some
1120 expansion items (e.g. "match"), thereby losing the authentication data.
1122 PH/21 Make -bV show the size of off_t variables so that the test suite can
1123 decide whether to run tests for quotas > 2G.
1125 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1126 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1127 filecount value is greater than 2G or if a quota value is greater than 2G
1128 on a system where the size of off_t is not greater than 4, a panic error
1131 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1132 never match. The debug and -bh output now contains an explicit error
1133 message indicating a malformed IPv4 address or mask.
1135 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1136 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1137 PH/23 above applies.
1139 PH/25 Do not write to syslog when running in the test harness. The only
1140 occasion when this arises is a failure to open the main or panic logs
1141 (for which there is an explicit test).
1143 PH/26 Added the /no_tell option to "control=freeze".
1145 PH/27 If a host name lookup failed very early in a connection, for example, if
1146 the IP address matched host_lookup and the reverse lookup yielded a name
1147 that did not have a forward lookup, an error message of the form "no IP
1148 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1149 could be logged. Now it outputs the IP address instead of "NULL".
1151 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1152 allows the sender and the authenticated sender to be set when
1153 submitting a message from within Exim. Since child_open_exim() is
1154 documented for local_scan(), the new function should be too.
1156 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1157 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1158 results in an empty string is now treated as unset.
1160 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1162 PH/31 Added sender_verify_fail logging option.
1164 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1165 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1166 tidied the source and removed it altogether.
1168 PH/33 When a queue run was abandoned because the load average was too high, a
1169 log line was always written; now it is written only if the queue_run log
1170 selector is set. In addition, the log line for abandonment now contains
1171 information about the queue run such as the pid. This is always present
1172 in "start" and "stop" lines but was omitted from the "abandon" line.
1174 PH/34 Omit spaces between a header name and the colon in the error message that
1175 is given when verify = headers_syntax fails (if there are lots of them,
1176 the message gets confusing).
1178 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1179 names, as there are now some PTR records that contain slashes. This check
1180 is only to protect against broken name servers that fall over on strange
1181 characters, so the fact that it applies to all lookups doesn't matter.
1183 PH/36 Now that the new test suite is complete, we can remove some of the
1184 special code in Exim that was needed for the old test suite. For example,
1185 sorting DNS records because real resolvers return them in an arbitrary
1186 order. The new test suite's fake resolver always returns records in the
1189 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1190 bounces) except when queue_only is set, to avoid logging races between
1191 the different processes.
1193 PH/38 Panic-die if .include specifies a non-absolute path.
1195 PH/39 A tweak to the "H" retry rule from its user.
1197 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1198 a label. They prevented compilation on older perls.
1200 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1201 a warning to be raised on newish perls.
1203 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1204 on queue. Changes to match documented behaviour of showing count of
1205 messages matching specified criteria.
1207 PH/40 Changed the default ident timeout from 30s to 5s.
1209 PH/41 Added support for the use of login_cap features, on those BSD systems
1210 that have them, for controlling the resources used by pipe deliveries.
1212 PH/42 The content-scanning code uses fopen() to create files in which to put
1213 message data. Previously it was not paying any attention to the mode of
1214 the files. Exim runs with umask(0) because the rest of the code creates
1215 files with open(), and sets the required mode explicitly. Thus, these
1216 files were ending up world-writeable. This was not a big issue, because,
1217 being within the spool directory, they were not world-accessible. I have
1218 created a function called modefopen, which takes an additional mode
1219 argument. It sets umask(777), creates the file, chmods it to the required
1220 mode, then resets the umask. All the relevant calls to fopen() in the
1221 content scanning code have been changed to use this function.
1223 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1224 to 24 hours. This avoids potential overflow problems when processing G
1225 and H retry rules. I suspect nobody ever tinkers with this value.
1227 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1229 PH/45 When the plaintext authenticator is running as a client, the server's
1230 challenges are checked to ensure they are valid base64 strings. By
1231 default, the authentication attempt is cancelled if an invalid string is
1232 received. Setting client_ignore_invalid_base64 true ignores these errors.
1233 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1234 they are received. Thus, the responses can be made to depend on the
1235 challenges. If an invalid string is ignored, an empty string is placed in
1238 PH/46 Messages that are created by the autoreply transport now contains a
1239 References: header, in accordance with RFCs 2822 and 3834.
1241 PH/47 Added authenticated_sender_force to the smtp transport.
1243 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1245 PH/49 Installed latest patch from the Sieve maintainer.
1247 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1248 was also set, the appendfile transport was unnecessarily scanning a
1249 directory of message files (e.g. for maildir delivery) to find the count
1250 of files (along with the size), even though it did not need this
1251 information. It now does the scan only if it needs to find either the
1252 size of the count of files.
1254 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1256 PH/52 Two bugs concerned with error handling when the smtp transport is
1259 (i) Exim was not creating retry information for temporary errors given
1260 for individual recipients after the DATA command when the smtp transport
1261 was used in LMTP mode. This meant that they could be retried too
1262 frequently, and not timed out correctly.
1264 (ii) Exim was setting the flag that allows error details to be returned
1265 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1266 recipients that were returned after the DATA command.
1268 PH/53 This is related to PH/52, but is more general: for any failing address,
1269 when detailed error information was permitted to be returned to the
1270 sender, but the error was temporary, then after the final timeout, only
1271 "retry timeout exceeded" was returned. Now it returns the full error as
1272 well as "retry timeout exceeded".
1274 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1275 do this, and (what is worse) MTAs that accept it.
1277 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1278 will now be deprecated.
1280 PH/56 New os.c-cygwin from the Cygwin maintainer.
1282 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1283 formats (previously only available in exim formats via -bpr, -bpru,
1284 and -bpra. Now also available in native and exiqgrep formats)
1286 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1287 with very large, slow to parse queues
1289 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1291 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1293 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1294 responses to authentication challenges, though it was showing the
1295 challenges; (ii) I've removed the CR characters from the debug output for
1298 PH/58 Allow for the insertion of a newline as well as a space when a string
1299 is turned into more than one encoded-word during RFC 2047 encoding. The
1300 Sieve code now uses this.
1302 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1303 data_4xx, lost_connection, tls_required.
1305 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1306 message was being sent as an SMTP response.
1308 PH/61 Add -l and -k options to exicyclog.
1310 PH/62 When verifying, if an address was redirected to one new address, so that
1311 verification continued, and the new address failed or deferred after
1312 having set something in $address_data, the value of $address_data was not
1313 passed back to the ACL. This was different to the case when no
1314 redirection occurred. The value is now passed back in both cases.
1316 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1317 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1318 use login_cap.h, so on its own it isn't the distinguishing feature. The
1319 new name refers directly to the setclassresources() function.
1321 PH/65 Added configuration files for NetBSD3.
1323 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1325 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1326 is preferred over IPv4.
1328 PH/68 The bounce_return_message and bounce_return_body options were not being
1329 honoured for bounces generated during the reception of non-SMTP messages.
1330 In particular, this applied to messages rejected by the ACL. This bug has
1331 been fixed. However, if bounce_return_message is true and bounce_return_
1332 body is false, the headers that are returned for a non-SMTP message
1333 include only those that have been read before the error was detected.
1334 (In the case of an ACL rejection, they have all been read.)
1336 PH/69 The HTML version of the specification is now built in a directory called
1337 spec_html instead of spec.html, because the latter looks like a path with
1338 a MIME-type, and this confuses some software.
1340 PH/70 Catch two compiler warnings in sieve.c.
1342 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1343 function verify_get_ident() calls ip_connect() to connect a socket, but
1344 if the "connect()" function timed out, ip_connect() used to close the
1345 socket. However, verify_get_ident() also closes the socket later, and in
1346 between Exim writes to the log, which may get opened at this point. When
1347 the socket was closed in ip_connect(), the log could get the same file
1348 descriptor number as the socket. This naturally causes chaos. The fix is
1349 not to close the socket in ip_connect(); the socket should be closed by
1350 the function that creates it. There was only one place in the code where
1351 this was missing, in the iplookup router, which I don't think anybody now
1352 uses, but I've fixed it anyway.
1354 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1355 well as to direct DNS lookups. Otherwise the handling of names in host
1356 lists is inconsistent and therefore confusing.
1362 PH/01 Two changes to the default runtime configuration:
1364 (1) Move the checks for relay_from_hosts and authenticated clients from
1365 after to before the (commented out) DNS black list checks.
1367 (2) Add control=submission to the relay_from_hosts and authenticated
1368 clients checks, on the grounds that messages accepted by these
1369 statements are most likely to be submissions.
1371 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1373 (1) Generate an error if the third argument for the ${prvs expansion is
1376 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1379 (3) Reset the variables that are obtained from the first argument of
1380 ${prvscheck and used in the second argument before leaving the code,
1381 because their memory is reclaimed, so using them afterwards may do
1384 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1385 one (it's much easier than Tom thought :-).
1387 (5) Because of (4), we can now allow for the use of $prvscheck_result
1388 inside the third argument.
1390 PH/03 For some reason, the default setting of PATH when running a command from
1391 a pipe transport was just "/usr/bin". I have changed it to
1394 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1395 anything to be listed in the output from -bV.
1397 PH/05 When a filter generated an autoreply, the entire To: header line was
1398 quoted in the delivery log line, like this:
1400 => >A.N.Other <ano@some.domain> <original@ddress> ...
1402 This has been changed so that it extracts the operative address. There
1403 may be more than one such address. If so, they are comma-separated, like
1406 => >ano@some.domain,ona@other.domain <original@ddress> ...
1408 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1409 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1410 not being looked up in the rDNS to get a host name, Exim was showing the
1411 IP address twice in Received: lines, even though the IP addresses were
1412 identical. For example:
1414 Received: from [1.2.3.4] (helo=[1.2.3.4])
1416 However, if the real host name was known, it was omitting the HELO data
1417 if it matched the actual IP address. This has been tidied up so that it
1418 doesn't show the same IP address twice.
1420 PH/07 When both +timestamp and +memory debugging was on, the value given by
1421 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1422 called by the debug printing, thereby overwriting the timestamp buffer.
1423 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1425 PH/08 When the original message was included in an autoreply transport, it
1426 always said "this is a copy of the message, including all the headers",
1427 even if body_only or headers_only was set. It now gives an appropriate
1430 PH/09 Applied a patch from the Sieve maintainer which:
1432 o fixes some comments
1433 o adds the (disabled) notify extension core
1434 o adds some debug output for the result of if/elsif tests
1435 o points to the current vacation draft in the documentation
1436 and documents the missing references header update
1440 o fixes a bug in processing the envelope test (when testing
1441 multiple envelope elements, the last element determinted the
1444 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1445 Electronic Mail") by including:
1447 Auto-submitted: auto-generated
1449 in the messages that it generates (bounce messages and others, such as
1450 warnings). In the case of bounce messages for non-SMTP mesages, there was
1451 also a typo: it was using "Auto_submitted" (underscore instead of
1452 hyphen). Since every message generated by Exim is necessarily in response
1453 to another message, thes have all been changed to:
1455 Auto-Submitted: auto-replied
1457 in accordance with these statements in the RFC:
1459 The auto-replied keyword:
1461 - SHOULD be used on messages sent in direct response to another
1462 message by an automatic process,
1464 - MUST NOT be used on manually-generated messages,
1466 - MAY be used on Delivery Status Notifications (DSNs) and Message
1467 Disposition Notifications (MDNs),
1469 - MUST NOT be used on messages generated by automatic or periodic
1470 processes, except for messages which are automatic responses to
1473 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1474 to the default Received: header definition.
1476 PH/12 Added log selector acl_warn_skipped (default on).
1478 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1479 variables because (a) they are in the wrong storage pool and (b) even if
1480 they were copied, it wouldn't work properly because of the caching.
1482 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1483 checking when decoding. Apparently there are clients that generate
1484 overlong encoded strings. Why am I not surprised?
1486 PH/15 If the first argument of "${if match_address" was not empty, but did not
1487 contain an "@" character, Exim crashed. Now it writes a panic log message
1488 and treats the condition as false.
1490 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1492 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1493 extension "envelope-auth". The code is finished and in agreement with
1494 other implementations, but there is no documentation so far and in fact,
1495 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1496 not changing the active code.
1498 Print executed "if" and "elsif" statements when debugging is used. This
1499 helps a great deal to understand what a filter does.
1501 Document more things not specified clearly in RFC3028. I had all this
1502 sorted out, when out of a sudden new issues came to my mind. Oops."
1504 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1507 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1508 canonical form (as documented). However, after a host name lookup from
1509 the IP address, check_host() was doing a simple string comparison with
1510 addresses acquired from the DNS when checking that the found name did
1511 have the original IP as one of its addresses. Since any found IPv6
1512 addresses are likely to be in abbreviated form, the comparison could
1513 fail. Luckily, there already exists a function for doing the comparison
1514 by converting both addresses to binary, so now that is used instead of
1515 the text comparison.
1517 PH/20 There was another similar case to PH/19, when a complete host name was
1518 given in a host list; looking up its IP address could give an abbreviated
1519 form, whereas the current host's name might or might not be abbreviated.
1520 The same fix has been applied.
1526 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1527 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1530 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1531 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1533 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1535 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1536 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1537 Also, for "Auto-Submitted", treat anything other than "no" as
1538 non-personal, in accordance with RFC 3834. (Previously it treated
1539 anything starting "auto-" as non-personal.)
1541 TF/01 The control=submission/name=... option had a problem with syntax
1542 errors if the name included a slash character. The /name= option
1543 now slurps the rest of the string, so it can include any characters
1544 but it must come last in the list of options (after /sender_retain
1547 PH/05 Some modifications to the interface to the fake nameserver for the new
1555 TK/01 Added the "success_on_redirect" address verification option. See
1556 NewStuff for rationale and an example.
1558 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1560 PH/02 Patch to exigrep to allow it to work on syslog lines.
1562 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1563 fread() to skip over the body file's header line, because in Cygwin the
1564 header line is locked and is inaccessible.
1566 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1567 co-exist for some time) to make it clear that it is the Exim ID that is
1568 referenced, not the Message-ID: header line.
1570 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1571 string_format() function, because snprintf() does not exist on all
1574 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1575 ${stat: expansion item.
1577 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1578 protocol synchronization error", to keep the pedants happy.
1580 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1581 well as for IRIX systems, when gcc is being used. See the host.c source
1584 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1586 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1589 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1590 transport and to the smtp transport in LMTP mode.
1592 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1594 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1595 run a filter in a subprocess. This could lead to confusion in subsequent
1596 lookups in the parent process. There should also be a search_tidyup() at
1597 the end of the subprocess.
1599 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1600 only if the host matched helo_try_verify_hosts, which caused the
1601 verification to occur when the EHLO/HELO command was issued. The ACL just
1602 tested the remembered result. Now, if a previous verification attempt has
1603 not happened, "verify = helo" does it there and then.
1605 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1607 TK/03 Fix log output including CR from clamd.
1609 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1610 exist provoked a memory error which could cause a segfault.
1612 PH/15 Installed PCRE 6.2
1614 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1616 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1617 of the problem. Specifically, suggested +O2 rather than +O1 for the
1620 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1622 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1623 record, the port was not correctly logged when the outgoing_port log
1624 selector was set (it logged the transort's default port).
1626 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1627 fallback_hosts, and "hosts" in the smtp transport.
1629 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1630 host errors such as "Connection refused".
1632 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1633 authentication with radiusclient 0.4.9:
1635 - Error returned from rc_read_config was caught wrongly
1636 - Username/password not passed on to radius server due to wrong length.
1638 The presumption is that some radiusclient API changes for 4.51/PH/17
1639 were not taken care of correctly. The code is still untested by me (my
1640 Linux distribution still has 0.3.2 of radiusclient), but it was
1641 contributed by a Radius user.
1643 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1644 expanding the "port" option of the smtp transport.
1646 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1647 while reading a MIME header. Thanks to Tom Hughes for a patch.
1649 PH/24 Include config.h inside local_scan.h so that configuration settings are
1652 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1653 that in an ACL for RCPT (for example), you can examine exactly what was
1656 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1657 commands, but it was not correctly comparing the address with the actual
1658 client host address. Thus, it would show the EHLO address in Received:
1659 header lines when this was not necessary.
1661 PH/27 Added the % operator to ${eval:}.
1663 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1664 it should be ignoring failures (because with -C, for example, it has lost
1665 privilege). It wasn't ignoring creation failures other than "already
1668 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1671 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1672 because an input line was too long, either on its own, or by virtue of
1673 too many continuations, the temporary file was not being removed, and the
1674 return code was incorrect.
1676 PH/31 Missing "BOOL" in function definition in filtertest.c.
1678 PH/32 Applied Sieve patches from the maintainer.
1680 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1682 PH/33 Added "verify = not_blind".
1684 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1685 Local/Makefile (with some defaults set). These are used in built scripts
1686 such as exicyclog, but they have never been used in the exim_install
1687 script (though there are many overriding facilities there). I have
1688 arranged that the exim_install script now takes note of these two
1691 PH/35 Installed configuration files for Dragonfly.
1693 PH/36 When a locally submitted message by a trusted user did not contain a
1694 From: header, and the sender address was obtained from -f or from an SMTP
1695 MAIL command, and the trusted user did not use -F to supply a sender
1696 name, $originator_name was incorrectly used when constructing a From:
1697 header. Furthermore, $originator_name was used for submission mode
1698 messages from external hosts without From: headers in a similar way,
1699 which is clearly wrong.
1701 PH/37 Added control=suppress_local_fixups.
1703 PH/38 When log_selector = +received_sender was set, and the addition of the
1704 sender made the log line's construction buffer exactly full, or one byte
1705 less than full, an overflow happened when the terminating "\n" was
1708 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1709 when the result of a list match is failure because a DNS lookup failed.
1711 PH/40 RM_COMMAND is now used in the building process.
1713 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1714 the "build-* directories that it finds.
1716 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1717 address in a domain literal was a prefix of an interface address.
1719 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1720 when verifying a sender address, unless rewrite_headers is false.
1722 PH/44 Wrote a long comment about why errors_to addresses are verified as
1723 recipients, not senders.
1725 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1726 the ratelimit ACL was added.
1728 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1730 PH/46 Added extra information about PostgreSQL errors to the error string.
1732 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1733 avoiding the need to install special zones in a real server. This is
1734 backwards compatible; if it can't find the fake resolver, it drops back.
1735 Thus, both old and new test suites can be run.
1737 TF/02 Added util/ratelimit.pl
1739 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1740 clock is set back in time.
1742 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1743 Candler <B.Candler@pobox.com>.
1745 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1746 OK for addresses that are the result of redirections.
1748 PH/48 A number of further additions for the benefit of the new test suite,
1749 including a fake gethostbyname() that interfaces to the fake DNS resolver
1752 TF/06 The fix for widen_domains has also been applied to qualify_single and
1753 search_parents which are the other dnslookup options that can cause
1756 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1759 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1761 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1762 service" if it encountered an MX record with an empty target hostname.
1763 The message is now "an MX or SRV record indicated no SMTP service".
1765 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1766 if the DNS of the sending site is misconfigured. This is quite a
1767 common situation. This change restores the behaviour of treating a
1768 helo verification defer as a failure.
1770 PH/51 If self=fail was set on a router, the bounce message did not include the
1771 actual error message.
1777 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1779 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1780 command itself ended in error, the underlying message about the transport
1781 filter timeout was being overwritten with the pipe command error. Now the
1782 underlying error message should be appended to the second error message.
1784 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1786 PH/02 Reduce the amount of output that "make" produces by default. Full output
1787 can still be requested.
1789 PH/03 The warning log line about a condition test deferring for a "warn" verb
1790 was being output only once per connection, rather than after each
1791 occurrence (because it was using the same function as for successful
1792 "warn" verbs). This seems wrong, so I have changed it.
1794 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1795 it should not have, which might have caused a crash in the right
1796 circumstances, but probably never did.
1798 PH/04 Installed a modified version of Tony Finch's patch to make submission
1799 mode fix the return path as well as the Sender: header line, and to
1800 add a /name= option so that you can make the user's friendly name appear
1803 TF/03 Added the control = fakedefer ACL modifier.
1805 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1806 Mark Lowes for thorough testing.
1808 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1810 TK/03 Merged latest SRS patch from Miles Wilton.
1812 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1813 with the definition in sysexits.h (which is #included earlier).
1814 Fortunately, Exim does not actually use EX_OK. The code used to try to
1815 preserve the sysexits.h value, by assumimg that macro definitions were
1816 scanned for macro replacements. I have been disabused of this notion,
1817 so now the code just undefines EX_OK before #including unistd.h.
1819 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1820 in the smtp transport. When a block could not be written in a single
1821 write() function, the timeout was being re-applied to each part-write.
1822 This seems wrong - if the receiver was accepting one byte at a time it
1823 would take for ever. The timeout is now adjusted when this happens. It
1824 doesn't have to be particularly precise.
1826 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1827 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1829 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1830 without a domain if the check to <postmaster@domain> fails.
1832 SC/01 Eximstats: added -xls and the ability to specify output files
1833 (patch written by Frank Heydlauf).
1835 SC/02 Eximstats: use FileHandles for outputing results.
1837 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1839 SC/04 Eximstats: fixed display of large numbers with -nvr option
1841 SC/05 Eximstats: fixed merging of reports with empty tables.
1843 SC/06 Eximstats: added the -include_original_destination flag
1845 SC/07 Eximstats: removed tabs and trailing whitespace.
1847 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1849 TK/06 MBOX spool code: Add real "From " MBOX separator line
1850 so the .eml file is really in mbox format (even though
1851 most programs do not really care). Patch from Alex Miller.
1853 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1854 The latter is generated from $received_to and is only set if the
1855 message has one envelope recipient. SA can use these headers,
1856 obviously out-of-the-box. Patch from Alex Miller.
1858 PH/08 The ${def test on a variable was returning false if the variable's
1859 value was "0", contrary to what the specification has always said!
1860 The result should be true unless the variable is empty.
1862 PH/09 The syntax error of a character other than { following "${if
1863 def:variable_name" (after optional whitespace) was not being diagnosed.
1864 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1865 accidental colon was present, for example, could give incorrect results.
1867 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1868 result is used (not including appendfile, where other changes are about
1871 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1872 This involved changing a lot of size variables from int to off_t. It
1873 should work with maildirs and everything.
1875 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1876 spamd dying while we are connected to it.
1878 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1881 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1882 function for BDB 4.3.
1884 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1886 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1887 the sources for PCRE have been re-arranged and more files are now
1890 PH/15 The code I had for printing potentially long long variables in PH/11
1891 above was not the best (it lost precision). The length of off_t variables
1892 is now inspected at build time, and an appropriate printing format (%ld
1893 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1894 to be "long long int" or "long int". This is needed for the internal
1895 formatting function string_vformat().
1897 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1898 the configuration file to be ":syslog", then the script "guesses" where
1899 the logs files are, rather than using the compiled in default. In our
1900 case the guess is not the same as the compiled default, so the script
1901 suddenly stopped working when I started to use syslog. The patch checks
1902 to see if log_file_path is "". If so, it attempts to read it from exim
1903 with no configuration file to get the compiled in version, before it
1904 falls back to the previous guessing code."
1906 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1907 implementing BATV in an Exim configuration. See NewStuff for the gory
1910 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1911 Makefile that are specific to HP-UX.
1913 PH/18 If the "use_postmaster" option was set for a recipient callout together
1914 with the "random" option, the postmaster address was used as the MAIL
1915 FROM address for the random test, but not for the subsequent recipient
1916 test. It is now used for both.
1918 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1919 patch removes a few documentation additions to RFC 3028, because the
1920 latest draft now contains them. It adds the new en;ascii-case comparator
1921 and a new error check for 8bit text in MIME parts. Comparator and
1922 require names are now matched exactly. I enabled the subaddress
1923 extension, but it is not well tested yet (read: it works for me)."
1925 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1926 rework some of the code of TK/09 above to avoid the hardwired use of
1927 "%lld" and "long long". Replaced the call to snprintf() with a call to
1930 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1931 records point to non-existent hosts", "retry timeout exceeded", and
1932 "retry time not reached for any host after a long failure period".
1934 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1935 experimental DomainKeys support:
1937 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1938 (2) On an error such as an illegally used "control", the wrong name for
1939 the control was given.
1941 These problems did NOT occur unless DomainKeys support was compiled.
1943 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1945 PH/24 Added ${if match_ip condition.
1947 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1948 databases so that it will be absolutely obvious if a crash occurs in the
1949 DB library. This is a regular occurrence (often caused by mis-matched
1952 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1953 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1954 picked up on a user's system that detects such things. There doesn't seem
1955 to be a gcc warning option for this - only an attribute that has to be
1956 put on the function's prototype. It seems that in Fedora Core 4 they have
1957 set this on a number of new functions. No doubt there will be more in due
1960 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1961 specify a transport. However, if an address that was verified by such a
1962 router was the subject of a callout, Exim crashed because it tried to
1963 read the rcpt_include_affixes from the non-existent transport. Now it
1964 just assumes that the setting of that option is false. This bug was
1965 introduced by 4.51/PH/31.
1967 PH/28 Changed -d+all to exclude +memory, because that information is very
1968 rarely of interest, but it makes the output a lot bigger. People tend to
1969 do -d+all out of habit.
1971 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1972 code in os-type was giving problems when libc.so lives in lib64, like on
1975 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1976 aren't the modern standard, and it seems that some systems' include files
1977 don't always have them. Exim was already checking for some of the newer
1978 ones like T_AAAA, and defining it itself. I've added checks for all the
1979 record types that Exim uses.
1981 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1982 not automatically generating a new one, as it is supposed to. This
1983 prevented TLS from working. If the file did exist, but contained invalid
1984 data, a new version was generated, as expected. It was only the case of a
1985 non-existent file that was broken.
1987 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1988 with a change in libdomainkeys > 0.64.
1990 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1991 from DNS. If the selector record carries the flag, it now has
1992 precedence over the domain-wide flag.
1994 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1996 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1997 the use of a port name that isn't defined in /etc/services) occurred, the
1998 message was deferred as in a normal delivery, and thus remained on the
1999 spool, instead of being failed because of the mua_wrapper setting. This
2000 is now fixed, and I tidied up some of the mua_wrapper messages at the
2003 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2004 the messages in a hash of arrays rather than using individual hashes.
2005 This is a bit cleaner and results in dramatic memory savings, albeit
2006 at a slight CPU cost.
2008 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2009 as requested by Marc Sherman.
2011 SC/10 Eximstats: added histograms for user specified patterns as requested
2014 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2016 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2017 fopen() in the content-scanning modules that did not already have it.
2023 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2024 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2026 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2028 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2030 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2031 to test Sieve filters that use "vacation".
2033 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2034 that changes the way the GnuTLS parameters are stored in the cache file.
2035 The new format can be generated externally. For backward compatibility,
2036 if the data in the cache doesn't make sense, Exim assumes it has read an
2037 old-format file, and it generates new data and writes a new file. This
2038 means that you can't go back to an older release without removing the
2041 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2042 work if there are any delivery delays because "one_time" forces the
2043 parent to be marked "delivered", so its unseen clone is never tried
2044 again. For this reason, Exim now forbids the simultaneous setting of
2047 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2048 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2049 entry if you want to know the details. The fix, however, neglected to
2050 consider the case when local delivery batching is involved. The test for
2051 "previously delivered" was not happening when checking to see if an
2052 address could be batched with a previous (undelivered) one; under
2053 certain circumstances this could lead to multiple deliveries to the same
2056 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2057 in its include files, and this causes problems building Exim.
2059 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2060 header_syntax) but Exim was just ignoring anything given after a slash.
2061 In particular, this caused confusion with an attempt to use "verify =
2062 reverse_host_lookup/defer_ok". An error is now given when options are
2063 supplied for verify items that do not have them. (Maybe reverse_host_
2064 lookup should have a defer_ok option, but that's a different point.)
2066 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2067 defined by RFC 821) to 2048, because there were problems with some AUTH
2068 commands, and RFC 1869 says the size should be increased for extended
2069 SMTP commands that take arguments.
2071 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2074 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2075 "unknown" error; now it says that the functionality isn't in the binary.
2077 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2078 an address' error message when a string expansion fails (syntax or
2079 whatever). Otherwise the password may appear in the log. Following change
2080 PH/42 below, there is no longer a chance of it appearing in a bounce
2083 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2085 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2086 of its addresses was ever tried. (Bugzilla bug #2.)
2088 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2089 the result incorrectly in the debug output. (It correctly added a newline
2090 to what was transported.)
2092 TF/01 Added $received_time.
2094 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2095 commented out examples of how to interface to a virus scanner and to
2096 SpamAssassin. Also added commented examples of av_scanner and
2097 spamd_address settings.
2099 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2100 and controls are allowed in which ACLs. There were a couple of minor
2101 errors. Some of the entries in the conditions table (which is a table of
2102 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2103 negation of where the condition IS allowed.
2105 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2107 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2108 header file does not have a version number, so I've had to invent a new
2109 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2110 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2111 radiusclient), but it was contributed by a Radius user.
2113 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2114 files or directories, for OpenSSL.
2116 PH/19 When an Exim process that is running as root has to create an Exim log
2117 file, it does so in a subprocess that runs as exim:exim so as to get the
2118 ownership right at creation (otherwise, other Exim processes might see
2119 the file with the wrong ownership). There was no test for failure of this
2120 fork() call, which would lead to the process getting stuck as it waited
2121 for a non-existent subprocess. Forks do occasionally fail when resources
2122 run out. I reviewed all the other calls to fork(); they all seem to check
2125 PH/20 When checking for unexpected SMTP input at connect time (before writing
2126 the banner), Exim was not dealing correctly with a non-positive return
2127 from the read() function. If the client had disconnected by this time,
2128 the result was a log entry for a synchronization error with an empty
2129 string after "input=" when read() returned zero. If read() returned -1
2130 (an event I could not check), uninitialized data bytes were printed.
2131 There were reports of junk text (parts of files, etc) appearing after
2134 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2136 PH/22 Added support for macro redefinition, and (re)definition in between
2137 driver and ACL definitions.
2139 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2140 forgetting to use the resulting value; it was using the unexpanded value.
2142 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2143 hadn't been configured. The fix is from Juergen Kreileder, who
2144 understands it better than I do:
2146 "Here's what I see happening with three configured cyrus_sasl
2147 authenticators configured (plain, login, cram-md5):
2149 On startup auth_cyrus_sasl_init() gets called for each of these.
2150 This means three calls to sasl_listmech() without a specified mech_list.
2151 => SASL tests which mechs of all available mechs actually work
2152 => three warnings about OTP not working
2153 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2155 With the patch, sasl_listmech() also gets called three times. But now
2156 SASL's mech_list option is set to the server_mech specified in the the
2157 authenticator. Or in other words, the answer from sasl_listmech()
2158 gets limited to just the mech you're testing for (which is different
2160 => the return list contains just 'plain' or 'login', 'cram-md5' or
2161 nothing depending on the value of ob->server_mech.
2163 I've just tested the patch: Authentication still works fine,
2164 unavailable mechs specified in the exim configuration are still
2165 caught, and the auth.log warnings about OTP are gone."
2167 PH/25 When debugging is enabled, the contents of the command line are added
2168 to the debugging output, even when log_selector=+arguments is not
2171 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2172 answer is "GNU", and only if the return is "GNU/something" is the answer
2175 PH/27 $acl_verify_message is now set immediately after the failure of a
2176 verification in an ACL, and so is available in subsequent modifiers. In
2177 particular, the message can be preserved by coding like this:
2179 warn !verify = sender
2180 set acl_m0 = $acl_verify_message
2182 Previously, $acl_verify_message was set only while expanding "message"
2183 and "log_message" when a very denied access.
2185 PH/28 Modified OS/os.c-Linux with
2187 -#ifndef OS_LOAD_AVERAGE
2188 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2190 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2191 nomenclature these days.)
2193 PH/29 Installed patch from the Sieve maintainer that adds the options
2194 sieve_useraddress and sieve_subaddress to the redirect router.
2196 PH/30 In these circumstances:
2197 . Two addresses routed to the same list of hosts;
2198 . First host does not offer TLS;
2199 . First host accepts first address;
2200 . First host gives temporary error to second address;
2201 . Second host offers TLS and a TLS session is established;
2202 . Second host accepts second address.
2203 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2204 and peerdn, if requested) that were in fact used only for the second
2207 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2208 attention to any local part prefix or suffix that was matched by the
2209 router that accepted the address. It now behaves in the same way as it
2210 does for delivery: the affixes are removed from the local part unless
2211 rcpt_include_affixes is set on the transport.
2213 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2214 timeout during the DATA phase of an incoming message.
2216 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2217 applied a patch sanctioned by the Sieve maintainer.
2219 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2220 the uid or gid is negative. A case of a negative gid caused this to be
2221 noticed. The fix allows for either to be negative.
2223 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2224 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2227 PH/36 The change PH/12 above was broken. Fixed it.
2229 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2230 the grounds that routing the same address twice would always produce the
2231 same answer. This might have been true once, but it is certainly no
2232 longer true now. Routing a child address may depend on the previous
2233 routing that produced that child. Some complicated redirection strategies
2234 went wrong when messages had multiple recipients, and made Exim's
2235 behaviour dependent on the order in which the addresses were given.
2237 I have moved the duplicate checking until after the routing is complete.
2238 Exim scans the addresses that are assigned to local and remote
2239 transports, and removes any duplicates. This means that more work will be
2240 done, as duplicates will always all be routed, but duplicates are
2241 presumably rare, so I don't expect this is of any significance.
2243 For deliveries to pipes, files, and autoreplies, the duplicate checking
2244 still happens during the routing process, since they are not going to be
2247 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2248 It corrects a timeout issue with spamd. This is Ian's comment: "The
2249 background is that sometimes spamd either never reads data from a
2250 connection it has accepted, or it never writes response data. The exiscan
2251 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2252 blindly assumes that writes won't block so it may never time out."
2254 PH/39 Allow G after quota size as well as K and M.
2256 PH/40 The value set for $authenticated_id in an authenticator may not contain
2257 binary zeroes or newlines because the value is written to log lines and
2258 to spool files. There was no check on this. Now the value is run through
2259 the string_printing() function so that such characters are converted to
2260 printable escape sequences.
2262 PH/41 $message_linecount is a new variable that contains the total number of
2263 lines in the message. Compare $body_linecount, which is the count for the
2266 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2267 bounce and delay warning messages, except in certain special cases, which
2270 (a) An SMTP error message from a remote host;
2271 (b) A message specified in a :fail: redirection;
2272 (c) A message specified in a "fail" command in a system filter;
2273 (d) A message specified in a FAIL return from the queryprogram router;
2274 (e) A message specified by the cannot_route_message router option.
2276 In these cases only, Exim does include the error details in bounce and
2277 warning messages. There are also a few cases where bland messages such
2278 as "unrouteable address" or "local delivery error" are given.
2280 PH/43 $value is now also set for the "else" part of a ${run expansion.
2282 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2283 being worked on, but at least Exim now implements the latest version to
2286 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2287 process to complete was treated as a delivery failure, a timeout while
2288 writing the message to the pipe was logged, but erroneously treated as a
2289 successful delivery. Such timeouts include transport filter timeouts. For
2290 consistency with the overall process timeout, these timeouts are now
2291 treated as errors, giving rise to delivery failures by default. However,
2292 there is now a new Boolean option for the pipe transport called
2293 timeout_defer, which, if set TRUE, converts the failures into defers for
2294 both kinds of timeout. A transport filter timeout is now identified in
2297 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2298 systems where "make" and "gmake" are different, calling "gmake" at top
2299 level broke things. I've arranged for the value of $(MAKE) to be passed
2300 from the Makefile to this script so that it can call the same version of
2304 A note about Exim versions 4.44 and 4.50
2305 ----------------------------------------
2307 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2308 changes of various kinds. As a consequence, a big documentation update was
2309 needed. This delayed the release for rather longer than seemed good, especially
2310 in the light of a couple of (minor) security issues. Therefore, the changes
2311 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2312 release. So 4.44 and 4.50 are in effect two different branches that both start
2315 I have left the 4.50 change log unchanged; it contains all the changes since
2316 4.43. The change log for 4.44 is below; many of its items are identical to
2317 those for 4.50. This seems to be the most sensible way to preserve the
2318 historical information.
2324 1. Minor wording change to the doc/README.SIEVE file.
2326 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2327 computation of the current number of files was incorrect.
2329 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2330 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2333 4. Give more explanation in the error message when the command for a transport
2334 filter fails to execute.
2336 5. There are several places where Exim runs a non-Exim command in a
2337 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2338 was being done only for the command run by the queryprogram router. It is
2339 now done for all such subprocesses. The other cases are: ${run, transport
2340 filters, and the commands run by the lmtp and pipe transports.
2342 6. Added CONFIGURE_GROUP build-time option.
2344 7. Some older OS have a limit of 256 on the maximum number of file
2345 descriptors. Exim was using setrlimit() to set 1000 as a large value
2346 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2347 systems. I've change it so that if it can't get 1000, it tries for 256.
2349 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2350 was an oversight, and furthermore, ever since the addition of extra
2351 controls (e.g. 4.43/32), the checks on when to allow different forms of
2352 "control" were broken. There should now be diagnostics for all cases when a
2353 control that does not make sense is encountered.
2355 9. Added the /retain_sender option to "control=submission".
2357 10. $recipients is now available in the predata ACL (oversight).
2359 11. Tidy the search cache before the fork to do a delivery from a message
2360 received from the command line. Otherwise the child will trigger a lookup
2361 failure and thereby defer the delivery if it tries to use (for example) a
2362 cached ldap connection that the parent has called unbind on.
2364 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2365 of $address_data from the recipient verification was clobbered by the
2366 sender verification.
2368 13. The value of address_data from a sender verification is now available in
2369 $sender_address_data in subsequent conditions in the ACL statement.
2371 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2373 15. Added a new option "connect=<time>" to callout options, to set a different
2376 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2377 was its contents. (It was OK if the option was not defined at all.)
2379 17. A "Completed" log line is now written for messages that are removed from
2380 the spool by the -Mrm option.
2382 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2383 information about exactly what failed.
2385 19. Added -dd to debug only the daemon process.
2387 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2388 handles timeouts, both on the server side and network timeouts. Renamed the
2389 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2391 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2392 It is now set to "smtps".
2394 22. $host_address is now set to the target address during the checking of
2395 ignore_target_hosts.
2397 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2398 being passed; this would have caused $sender_host_name to have been used if
2399 matching the list had actually called for a host name (not very likely,
2400 since this list is usually IP addresses). A host name is now passed as
2403 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2404 code that specifies a non-restarting handler (typically sigaction() in
2405 modern systems) in an attempt to fix a rare and obscure crash bug.
2407 25. Narrowed the window for a race in the daemon that could cause it to ignore
2408 SIGCHLD signals. This is not a major problem, because they are used only to
2409 wake it up if nothing else does.
2411 26. A malformed maildirsize file could cause Exim to calculate negative values
2412 for the mailbox size or file count. Odd effects could occur as a result.
2413 The maildirsize information is now recalculated if the size or filecount
2416 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2417 support for a long time. Removed HAVE_SYS_VFS_H.
2419 28. Installed the latest version of exipick from John Jetmore.
2421 29. In an address list, if the pattern was not a regular expression, an empty
2422 subject address (from a bounce message) matched only if the pattern was an
2423 empty string. Non-empty patterns were not even tested. This was the wrong
2424 because it is perfectly reasonable to use an empty address as part of a
2425 database query. An empty address is now tested by patterns that are
2426 lookups. However, all the other forms of pattern expect the subject to
2427 contain a local part and a domain, and therefore, for them, an empty
2428 address still always fails if the pattern is not itself empty.
2430 30. Exim went into a mad DNS loop when attempting to do a callout where the
2431 host was specified on an smtp transport, and looking it up yielded more
2432 than one IP address.
2434 31. Re-factored the code for checking spool and log partition space into a
2435 function that finds that data and another that does the check. The former
2436 is then used to implement four new variables: $spool_space, $log_space,
2437 $spool_inodes, and $log_inodes.
2439 32. The RFC2047 encoding function was originally intended for short strings
2440 such as real names; it was not keeping to the 75-character limit for
2441 encoded words that the RFC imposes. It now respects the limit, and
2442 generates multiple encoded words if necessary. To be on the safe side, I
2443 have increased the buffer size for the ${rfc2047: expansion operator from
2446 33. It is now permitted to omit both strings after an "if" condition; if the
2447 condition is true, the result is "true". As before, when the second string
2448 is omitted, a false condition yields an empty string. This makes it less
2449 cumbersome to write custom ACL and router conditions.
2451 34. Failure to deliver a bounce message always caused it to be frozen, even if
2452 there was an errors_to setting on the router. The errors_to setting is now
2455 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2456 canonical form (fully expanded) before being placed in
2457 $sender_host_address.
2459 36. The table in the code that translates DNS record types into text (T_A to
2460 "A" for instance) was missing entries for NS and CNAME. It is just possible
2461 that this could have caused confusion if both these types were looked up
2462 for the same domain, because the text type is used as part of Exim's
2463 per-process caching. But the chance of anyone hitting this buglet seems
2466 37. The dnsdb lookup has been extended in a number of ways.
2468 (1) There is a new type, "zns", which walks up the domain tree until it
2469 finds some nameserver records. It should be used with care.
2471 (2) There is a new type, "mxh", which is like "mx" except that it returns
2472 just the host names, not the priorities.
2474 (3) It is now possible to give a list of domains (or IP addresses) to be
2475 looked up. The behaviour when one of the lookups defers can be
2476 controlled by a keyword.
2478 (4) It is now possible to specify the separator character for use when
2479 multiple records are returned.
2481 38. The dnslists ACL condition has been extended: it is now possible to supply
2482 a list of IP addresses and/or domains to be looked up in a particular DNS
2485 39. Added log_selector=+queue_time_overall.
2487 40. When running the queue in the test harness, wait just a tad after forking a
2488 delivery process, to get repeatability of debugging output.
2490 41. Include certificate and key file names in error message when GnuTLS fails
2491 to set them up, because the GnuTLS error message doesn't include the name
2492 of the failing file when there is a problem reading it.
2494 42. Allow both -bf and -bF in the same test run.
2496 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2498 44. The "Exiscan patch" is now merged into the mainline Exim source.
2500 45. Sometimes the final signoff response after QUIT could fail to get
2501 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2502 < 0 before doing a fflush(). This bug looks as though it goes back to the
2503 introduction of TLS in release 3.20, but "sometimes" must have been rare
2504 because the tests only now provoked it.
2506 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2507 (this can affect the format of dates).
2509 47. exim_tidydb, when checking for the continued existence of a message for
2510 which it has found a message-specific retry record, was not finding
2511 messages that were in split spool directories. Consequently, it was
2512 deleting retry records that should have stayed in existence.
2514 48. Steve fixed some bugs in eximstats.
2516 49. The SPA authentication driver was not abandoning authentication and moving
2517 on to the next authenticator when an expansion was forced to fail,
2518 contradicting the general specification for all authenticators. Instead it
2519 was generating a temporary error. It now behaves as specified.
2521 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2522 (the order specifies the preference for clients). The order is now AES256,
2523 AES128, 3DES, ARCFOUR128.
2525 51. Small patch to Sieve code - explicitly set From: when generating an
2528 52. Exim crashed if a remote delivery caused a very long error message to be
2529 recorded - for instance if somebody sent an entire SpamAssassin report back
2530 as a large number of 550 error lines. This bug was coincidentally fixed by
2531 increasing the size of one of Exim's internal buffers (big_buffer) that
2532 happened as part of the Exiscan merge. However, to be on the safe side, I
2533 have made the code more robust (and fixed the comments that describe what
2536 53. Now that there can be additional text after "Completed" in log lines (if
2537 the queue_time_overall log selector is set), a one-byte patch to exigrep
2538 was needed to allow it to recognize "Completed" as not the last thing in
2541 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2542 patch that reportedly fixes this has been added. I am not expert enough to
2543 create a test for it. This is what the patch creator wrote:
2545 "I found a little strange behaviour of ldap code when working with
2546 Windows 2003 AD Domain, where users was placed in more than one
2547 Organization Units. When I tried to give exim partial DN, the exit code
2548 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2549 But simultaneously result of request was absolutely normal ldap result,
2550 so I produce this patch..."
2552 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2553 REFERENCE, so I have modified the code to exclude the patch when that macro
2556 55. Some experimental protocols are using DNS PTR records for new purposes. The
2557 keys for these records are domain names, not reversed IP addresses. The
2558 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2559 leaves it alone. Component reversal etc. now happens only for IP addresses.
2562 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2564 57. Double the size of the debug message buffer (to 2048) so that more of very
2565 long debug lines gets shown.
2567 58. The exicyclog utility now does better if the number of log files to keep
2568 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2570 59. Two changes related to the smtp_active_hostname option:
2572 (1) $smtp_active_hostname is now available as a variable.
2573 (2) The default for smtp_banner uses $smtp_active_hostname instead
2574 of $primary_hostname.
2576 60. The host_aton() function is supposed to be passed a string that is known
2577 to be a valid IP address. However, in the case of IPv6 addresses, it was
2578 not checking this. This is a hostage to fortune. Exim now panics and dies
2579 if the condition is not met. A case was found where this could be provoked
2580 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2581 components; fortuitously, this particular loophole had already been fixed
2582 by change 4.50/55 above.
2584 If there are any other similar loopholes, the new check in host_aton()
2585 itself should stop them being exploited. The report I received stated that
2586 data on the command line could provoke the exploit when Exim was running as
2587 exim, but did not say which command line option was involved. All I could
2588 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2589 running as the user.
2592 61. There was a buffer overflow vulnerability in the SPA authentication code
2593 (which came originally from the Samba project). I have added a test to the
2594 spa_base64_to_bits() function which I hope fixes it.
2597 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2598 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2600 63. The daemon start-up calls getloadavg() while still root for those OS that
2601 need the first call to be done as root, but it missed one case: when
2602 deliver_queue_load_max is set with deliver_drop_privilege. This is
2603 necessary for the benefit of the queue runner, because there is no re-exec
2604 when deliver_drop_privilege is set.
2606 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2607 This has been fixed.
2609 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2610 was in use, was not putting the data itself into the right store pool;
2611 consequently, it could be overwritten for a subsequent message in the same
2612 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2615 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2617 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2618 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2619 the function were treating the return as a boolean value, which happened to
2620 work because 0=false and not-0=true, but is not correct code.
2622 68. The host_aton() function was not handling scoped IPv6 addresses (those
2623 with, for example, "%eth0" on the end) correctly.
2625 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2626 negated items (that is, ~something) in unsigned ints. Some compilers
2627 apparently mutter when there is no cast.
2629 70. If an address verification called from an ACL failed, and did not produce a
2630 user-specific message (i.e. there was only a "system" message), nothing was
2631 put in $acl_verify_message. In this situation, it now puts the system
2634 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2635 session; change 4.31/43 added the unwanted input to the log line - except
2636 that it did not do this in the start of session case. It now does.
2638 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2639 This is wrong and can cause the other end to generate a synchronization
2640 error if it is another Exim or anything else that does the synchronization
2641 check. A QUIT command is no longer sent after a timeout.
2643 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2644 during host lookups.
2646 74. The defer_ok option of callout verification was not working if it was used
2647 when verifying addresses in header lines, that is, for this case:
2649 verify = header_sender/callout=defer_ok
2651 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2652 those file descriptors could be used for SMTP connections. If anything
2653 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2654 could be sent to the SMTP client, causing chaos. The daemon now opens
2655 stdin, stdout, and stderr to /dev/null when it puts itself into the
2658 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2659 log by default. The user can override this with suitable Perl magic.
2661 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2662 the log message when discard triggers, was not working for the DATA ACL or
2663 for the non-SMTP ACL.
2665 78. Error message wording change in sieve.c.
2667 79. If smtp_accept_max_per_host was set, the number of connections could be
2668 restricted to fewer than expected, because the daemon was trying to set up
2669 a new connection before checking whether the processes handling previous
2670 connections had finished. The check for completed processes is now done
2671 earlier. On busy systems, this bug wouldn't be noticed because something
2672 else would have woken the daemon, and it would have reaped the completed
2675 80. If a message was submitted locally by a user whose login name contained one
2676 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2677 It caused a spool format error. I have fixed the spool reading code. A
2678 related problem was that the "from" clause in the Received: line became
2679 illegal because of the space(s). It is now covered by ${quote_local_part.
2681 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2684 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2686 83. Patch from Sieve maintainer to fix unterminated string problem in
2687 "vacation" handling.
2689 84. Some minor changes to the Linux configuration files to help with other
2690 OS variants using glibc.
2692 85. One more patch for Sieve to update vacation handling to latest spec.
2695 ----------------------------------------------------
2696 See the note above about the 4.44 and 4.50 releases.
2697 ----------------------------------------------------
2703 1. Change 4.43/35 introduced a bug that caused file counts to be
2704 incorrectly computed when quota_filecount was set in an appendfile
2707 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2708 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2711 3. Give more explanation in the error message when the command for a transport
2712 filter fails to execute.
2714 4. There are several places where Exim runs a non-Exim command in a
2715 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2716 was being done only for the command run by the queryprogram router. It is
2717 now done for all such subprocesses. The other cases are: ${run, transport
2718 filters, and the commands run by the lmtp and pipe transports.
2720 5. Some older OS have a limit of 256 on the maximum number of file
2721 descriptors. Exim was using setrlimit() to set 1000 as a large value
2722 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2723 systems. I've change it so that if it can't get 1000, it tries for 256.
2725 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2726 was an oversight, and furthermore, ever since the addition of extra
2727 controls (e.g. 4.43/32), the checks on when to allow different forms of
2728 "control" were broken. There should now be diagnostics for all cases when a
2729 control that does not make sense is encountered.
2731 7. $recipients is now available in the predata ACL (oversight).
2733 8. Tidy the search cache before the fork to do a delivery from a message
2734 received from the command line. Otherwise the child will trigger a lookup
2735 failure and thereby defer the delivery if it tries to use (for example) a
2736 cached ldap connection that the parent has called unbind on.
2738 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2739 of $address_data from the recipient verification was clobbered by the
2740 sender verification.
2742 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2743 was its contents. (It was OK if the option was not defined at all.)
2745 11. A "Completed" log line is now written for messages that are removed from
2746 the spool by the -Mrm option.
2748 12. $host_address is now set to the target address during the checking of
2749 ignore_target_hosts.
2751 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2752 being passed; this would have caused $sender_host_name to have been used if
2753 matching the list had actually called for a host name (not very likely,
2754 since this list is usually IP addresses). A host name is now passed as
2757 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2758 code that specifies a non-restarting handler (typically sigaction() in
2759 modern systems) in an attempt to fix a rare and obscure crash bug.
2761 15. Narrowed the window for a race in the daemon that could cause it to ignore
2762 SIGCHLD signals. This is not a major problem, because they are used only to
2763 wake it up if nothing else does.
2765 16. A malformed maildirsize file could cause Exim to calculate negative values
2766 for the mailbox size or file count. Odd effects could occur as a result.
2767 The maildirsize information is now recalculated if the size or filecount
2770 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2771 support for a long time. Removed HAVE_SYS_VFS_H.
2773 18. Updated exipick to current release from John Jetmore.
2775 19. Allow an empty sender to be matched against a lookup in an address list.
2776 Previously the only cases considered were a regular expression, or an
2779 20. Exim went into a mad DNS lookup loop when doing a callout where the
2780 host was specified on the transport, if the DNS lookup yielded more than
2783 21. The RFC2047 encoding function was originally intended for short strings
2784 such as real names; it was not keeping to the 75-character limit for
2785 encoded words that the RFC imposes. It now respects the limit, and
2786 generates multiple encoded words if necessary. To be on the safe side, I
2787 have increased the buffer size for the ${rfc2047: expansion operator from
2790 22. Failure to deliver a bounce message always caused it to be frozen, even if
2791 there was an errors_to setting on the router. The errors_to setting is now
2794 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2795 canonical form (fully expanded) before being placed in
2796 $sender_host_address.
2798 24. Updated eximstats to version 1.33
2800 25. Include certificate and key file names in error message when GnuTLS fails
2801 to set them up, because the GnuTLS error message doesn't include the name
2802 of the failing file when there is a problem reading it.
2804 26. Expand error message when OpenSSL has problems setting up cert/key files.
2807 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2808 (this can affect the format of dates).
2810 28. exim_tidydb, when checking for the continued existence of a message for
2811 which it has found a message-specific retry record, was not finding
2812 messages that were in split spool directories. Consequently, it was
2813 deleting retry records that should have stayed in existence.
2815 29. eximstats updated to version 1.35
2816 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2817 1.35 - bugfix such that pie charts by volume are generated correctly
2819 30. The SPA authentication driver was not abandoning authentication and moving
2820 on to the next authenticator when an expansion was forced to fail,
2821 contradicting the general specification for all authenticators. Instead it
2822 was generating a temporary error. It now behaves as specified.
2824 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2825 (the order specifies the preference for clients). The order is now AES256,
2826 AES128, 3DES, ARCFOUR128.
2828 31. Small patch to Sieve code - explicitly set From: when generating an
2831 32. Exim crashed if a remote delivery caused a very long error message to be
2832 recorded - for instance if somebody sent an entire SpamAssassin report back
2833 as a large number of 550 error lines. This bug was coincidentally fixed by
2834 increasing the size of one of Exim's internal buffers (big_buffer) that
2835 happened as part of the Exiscan merge. However, to be on the safe side, I
2836 have made the code more robust (and fixed the comments that describe what
2839 33. Some experimental protocols are using DNS PTR records for new purposes. The
2840 keys for these records are domain names, not reversed IP addresses. The
2841 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2842 leaves it alone. Component reversal etc. now happens only for IP addresses.
2845 34. The host_aton() function is supposed to be passed a string that is known
2846 to be a valid IP address. However, in the case of IPv6 addresses, it was
2847 not checking this. This is a hostage to fortune. Exim now panics and dies
2848 if the condition is not met. A case was found where this could be provoked
2849 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2850 components; fortuitously, this particular loophole had already been fixed
2851 by change 4.50/55 or 4.44/33 above.
2853 If there are any other similar loopholes, the new check in host_aton()
2854 itself should stop them being exploited. The report I received stated that
2855 data on the command line could provoke the exploit when Exim was running as
2856 exim, but did not say which command line option was involved. All I could
2857 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2858 running as the user.
2861 35. There was a buffer overflow vulnerability in the SPA authentication code
2862 (which came originally from the Samba project). I have added a test to the
2863 spa_base64_to_bits() function which I hope fixes it.
2866 36. The daemon start-up calls getloadavg() while still root for those OS that
2867 need the first call to be done as root, but it missed one case: when
2868 deliver_queue_load_max is set with deliver_drop_privilege. This is
2869 necessary for the benefit of the queue runner, because there is no re-exec
2870 when deliver_drop_privilege is set.
2872 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2873 was in use, was not putting the data itself into the right store pool;
2874 consequently, it could be overwritten for a subsequent message in the same
2875 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2878 38. Sometimes the final signoff response after QUIT could fail to get
2879 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2880 < 0 before doing a fflush(). This bug looks as though it goes back to the
2881 introduction of TLS in release 3.20, but "sometimes" must have been rare
2882 because the tests only now provoked it.
2888 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2889 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2890 Now it returns an integer. A number of calls were still expecting a T/F
2891 return. Fortuitously, in all cases, the tests worked in OK situations,
2892 which is the norm. However, things would have gone wrong on any write
2893 failures on the smtp file descriptor. This function is used when sending
2894 messages over SMTP and also when doing verify callouts.
2896 2. When Exim is called to do synchronous delivery of a locally submitted
2897 message (the -odf or -odi options), it no longer closes stderr before doing
2900 3. Implemented the mua_wrapper option.
2902 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2904 5. Implemented the functions header_remove(), header_testname(),
2905 header_add_at_position(), and receive_remove_recipient(), and exported them
2908 6. If an ACL "warn" statement specified the addition of headers, Exim already
2909 inserted X-ACL-Warn: at the start if there was no header name. However, it
2910 was not making this test for the second and subsequent header lines if
2911 there were newlines in the string. This meant that an invalid header could
2912 be inserted if Exim was badly configured.
2914 7. Allow an ACL "warn" statement to add header lines at the start or after all
2915 the Received: headers, as well as at the end.
2917 8. Added the rcpt_4xx retry error code.
2919 9. Added postmaster_mailfrom=xxx to callout verification option.
2921 10. Added mailfrom=xxxx to the callout verification option, for verify=
2924 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2925 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2927 12. Inserted some casts to stop certain compilers warning when using pointer
2928 differences as field lengths or precisions in printf-type calls (mostly
2929 affecting debugging statements).
2931 13. Added optional readline() support for -be (dynamically loaded).
2933 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2934 same clock tick as a message's arrival, so that its received time was the
2935 same as the "first fail" time on the retry record, and that message
2936 remained on the queue past the ultimate address timeout, every queue runner
2937 would try a delivery (because it was past the ultimate address timeout) but
2938 after another failure, the ultimate address timeout, which should have then
2939 bounced the address, did not kick in. This was a "< instead of <=" error;
2940 in most cases the first failure would have been in the next clock tick
2941 after the received time, and all would be well.
2943 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2944 being recognized when the domain list was tested by the match_domain
2945 condition in an expansion string.
2947 16. Added the ${str2b64: operator.
2949 17. Exim was always calling setrlimit() to set a large limit for the number of
2950 processes, without checking whether the existing limit was already
2951 adequate. (It did check for the limit on file descriptors.) Furthermore,
2952 errors from getrlimit() and setrlimit() were being ignored. Now they are
2953 logged to the main and panic logs, but Exim does carry on, to try to do its
2954 job under whatever limits there are.
2956 18. Imported PCRE 5.0.
2958 19. Trivial typo in log message " temporarily refused connection" (the leading
2961 20. If the log selector return_path_on_delivery was set and an address was
2962 redirected to /dev/null, the delivery process crashed because it assumed
2963 that a return path would always be set for a "successful" delivery. In this
2964 case, the whole delivery is bypassed as an optimization, and therefore no
2967 21. Internal re-arrangement: the function for sending a challenge and reading
2968 a response while authentication was assuming a zero-terminated challenge
2969 string. It's now changed to take a pointer and a length, to allow for
2970 binary data in such strings.
2972 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2974 23. Exim was not respecting finduser_retries when seeking the login of the
2975 uid under which it was called; it was always trying 10 times. (The default
2976 setting of finduser_retries is zero.) Also, it was sleeping after the final
2977 failure, which is pointless.
2979 24. Implemented tls_on_connect_ports.
2981 25. Implemented acl_smtp_predata.
2983 26. If the domain in control=submission is set empty, Exim assumes that the
2984 authenticated id is a complete email address when it generates From: or
2985 Sender: header lines.
2987 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2988 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2989 chown and chgrp in /bin and hostname in /usr/bin.
2991 28. Exim was keeping the "process log" file open after each use, just as it
2992 does for the main log. This opens the possibility of it remaining open for
2993 long periods when the USR1 signal hits a daemon. Occasional processlog
2994 errors were reported, that could have been caused by this. Anyway, it seems
2995 much more sensible not to leave this file open at all, so that is what now
2998 29. The long-running daemon process does not normally write to the log once it
2999 has entered its main loop, and it closes the log before doing so. This is
3000 so that log files can straightforwardly be renamed and moved. However,
3001 there are a couple of unusual error situations where the daemon does write
3002 log entries, and I had neglected to close the log afterwards.
3004 30. The text of an SMTP error response that was received during a remote
3005 delivery was being truncated at 512 bytes. This is too short for some of
3006 the long messages that one sometimes sees. I've increased the limit to
3009 31. It is now possible to make retry rules that apply only when a message has a
3010 specific sender, in particular, an empty sender.
3012 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3013 it possible to be selective about when SMTP synchronization is enforced.
3015 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3017 32. Implemented hosts_connection_nolog.
3019 33. Added an ACL for QUIT.
3021 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3024 35. Added mailbox_size and mailbox_filecount to appendfile.
3026 36. Added control = no_multiline_responses to ACLs.
3028 37. There was a bug in the logic of the code that waits for the clock to tick
3029 in the case where the clock went backwards by a substantial amount such
3030 that the microsecond fraction of "now" was more than the microsecond
3031 fraction of "then" (but the whole seconds number was less).
3033 38. Added support for the libradius Radius client library this is found on
3034 FreeBSD (previously only the radiusclient library was supported).
3040 1. When certain lookups returned multiple values in the form name=value, the
3041 quoting of the values was not always being done properly. Specifically:
3042 (a) If the value started with a double quote, but contained no whitespace,
3044 (b) If the value contained whitespace other than a space character (i.e.
3045 tabs or newlines or carriage returns) it was not quoted.
3046 This fix has been applied to the mysql and pgsql lookups by writing a
3047 separate quoting function and calling it from the lookup code. The fix
3048 should probably also be applied to nisplus, ibase and oracle lookups, but
3049 since I cannot test any of those, I have not disturbed their existing code.
3051 2. A hit in the callout cache for a specific address caused a log line with no
3052 reason for rejecting RCPT. Now it says "Previous (cached) callout
3053 verification failure".
3055 3. There was an off-by-one bug in the queryprogram router. An over-long
3056 return line was truncated at 256 instead of 255 characters, thereby
3057 overflowing its buffer with the terminating zero. As well as fixing this, I
3058 have increased the buffer size to 1024 (and made a note to document this).
3060 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3061 when Exim is waiting for an SMTP response from a remote server, Exim
3062 restarts its select() call on the socket, thereby resetting its timeout.
3063 This is not a problem when such interrupts are rare. Somebody set up a cron
3064 job to run exiwhat every 2 minutes, which is less than the normal select()
3065 timeout (5 or 10 minutes). This meant that the select() timeout never
3066 kicked in because it was always reset. I have fixed this by comparing the
3067 time when an interrupt arrives with the time at the start of the first call
3068 to select(). If more time than the timeout has elapsed, the interrupt is
3069 treated as a timeout.
3071 5. Some internal re-factoring in preparation for the addition of Sieve
3072 extensions (by MH). In particular, the "personal" test is moved to a
3073 separate function, and given an option for scanning Cc: and Bcc: (which is
3074 not set for Exim filters).
3076 6. When Exim created an email address using the login of the caller as the
3077 local part (e.g. when creating a From: or Sender: header line), it was not
3078 quoting the local part when it contained special characters such as @.
3080 7. Installed new OpenBSD configuration files.
3082 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3083 try to make them clearer.
3085 9. Callout options, other than the timeout value, were being ignored when
3086 verifying sender addresses in header lines. For example, when using
3088 verify = header_sender/callout=no_cache
3090 the cache was (incorrectly) being used.
3092 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3093 only the Cygwin environment.
3095 11. When return_path_on_delivery was set as a log selector, if different remote
3096 addresses in the same message used different return paths and parallel
3097 remote delivery occurred, the wrong values would sometimes be logged.
3098 (Whenever a remote delivery process finished, the return path value from
3099 the most recently started remote delivery process was logged.)
3101 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3102 lines when AUTH and/or TLS are in use. This is the "received protocol"
3103 field. Exim used to use "asmtp" for authenticated SMTP, without any
3104 indication (in the protocol name) for TLS use. Now it follows the RFC and
3105 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3106 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3107 names appear in log lines as well as in Received: header lines.
3109 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3110 extensions, and comparison tests, and to fix some bugs.
3112 14. Changes to the "personal" filter test:
3114 (1) The test was buggy in that it was just doing the equivalent of
3115 "contains" tests on header lines. For example, if a user's address was
3116 anne@some.where, the "personal" test would incorrectly be true for
3118 To: susanne@some.where
3120 This test is now done by extracting each address from the header in turn,
3121 and checking the entire address. Other tests that are part of "personal"
3122 are now done using regular expressions (for example, to check local parts
3123 of addresses in From: header lines).
3125 (2) The list of non-personal local parts in From: addresses has been
3126 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3127 taken from the Sieve specification recommendations.
3129 (3) If the message contains any header line starting with "List-" it is
3130 treated as non-personal.
3132 (4) The test for "circular" in the Subject: header line has been removed
3133 because it now seems ill-conceived.
3135 15. Minor typos in src/EDITME comments corrected.
3137 16. Installed latest exipick from John Jetmore.
3139 17. If headers_add on a router specified a text string that was too long for
3140 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3141 of string_sprintf() is now avoided.
3143 18. $message_body_size was not set (it was always zero) when running the DATA
3144 ACL and the local_scan() function.
3146 19. For the "mail" command in an Exim filter, no default was being set for
3147 the once_repeat time, causing a random time value to be used if "once" was
3148 specified. (If the value happened to be <= 0, no repeat happened.) The
3149 default is now 0s, meaning "never repeat". The "vacation" command was OK
3150 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3151 (I found it when inspecting the code).
3153 20. There is now an overall timeout for performing a callout verification. It
3154 defaults to 4 times the callout timeout, which applies to individual SMTP
3155 commands during the callout. The overall timeout applies when there is more
3156 than one host that can be tried. The timeout is checked before trying the
3157 next host. This prevents very long delays if there are a large number of
3158 hosts and all are timing out (e.g. when the network connections are timing
3159 out). The value of the overall timeout can be changed by specifying an
3160 additional sub-option for "callout", called "maxwait". For example:
3162 verify = sender/callout=5s,maxwait=20s
3164 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3165 to the end before writing, but this should make it even safer).
3167 22. Exim was forgetting that it had advertised PIPELINING for the second and
3168 subsequent messages on an SMTP connection. It was also not resetting its
3169 memory on STARTTLS and an internal HELO.
3171 23. When Exim logs an SMTP synchronization error within a session, it now
3172 records whether PIPELINING has been advertised or not.
3174 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3175 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3176 rather than long int.
3178 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3180 26. Added the never_mail option to autoreply.
3186 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3187 crash if the getsockname() call failed; this can happen if a connection is
3188 closed very soon after it is established. The problem was simply in the
3189 order in which certain operations were done, causing Exim to try to write
3190 to the SMTP stream before it had set up the file descriptor. The bug has
3191 been fixed by making things happen in the correct order.
3197 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3198 before the connection was closed, thus losing the rejection response.
3200 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3201 some early Solaris releases, but causes trouble in current releases where
3202 socklen_t is defined.
3204 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3207 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3210 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3211 that releasing the top of it at the end releases what was used for sub-
3212 expansions (unless the block got too big). However, discard this block if
3213 the first thing is a variable or header, so that we can use its block when
3214 it is dynamic (useful for very large $message_headers, for example).
3216 6. Lookups now cache *every* query, not just the most recent. A new, separate
3217 store pool is used for this. It can be recovered when all lookup caches are
3218 flushed. Lookups now release memory at the end of their result strings.
3219 This has involved some general refactoring of the lookup sources.
3221 7. Some code has been added to the store_xxx() functions to reduce the amount
3222 of flapping under certain conditions.
3224 8. log_incoming_interface used to affect only the <= reception log lines. Now
3225 it causes the local interface and port to be added to several more SMTP log
3226 lines, for example "SMTP connection from", and rejection lines.
3228 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3230 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3232 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3233 could be overwritten at the end of the current message (or the start of a
3234 new message if it was set in a HELO ACL). The value is now preserved for
3235 the duration of the SMTP connection.
3237 12. If a transport had a headers_rewrite setting, and a matching header line
3238 contained an unqualified address, that address was qualified, even if it
3239 did not match any rewriting rules. The underlying bug was that the values
3240 of the flags that permit the existence of unqualified sender and recipient
3241 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3242 non-local messages, and by -bnq for local messages) were not being
3243 preserved with the message after it was received.
3245 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3246 "next input=" as part of the text comprising the host identity instead of
3247 the correct text. The code was using the same buffer for two different
3248 strings. However, depending on which order the printing function evaluated
3249 its arguments, the bug did not always show up. Under Linux, for example, my
3250 test suite worked just fine.
3252 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3253 doesn't work with some older versions of Perl. It has been changed to "my",
3254 which in any case is probably the better facility to use.
3256 15. A really picky compiler found some instances of statements for creating
3257 error messages that either had too many or two few arguments for the format
3260 16. The size of the buffer for calls to the DNS resolver has been increased
3261 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3262 for addresses that have a lot of PTR records. This alleviates a problem; it
3263 does not fully solve it.
3265 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3266 buffer now truncates the list and logs the incident, which is the same
3267 action as happens when Exim is looking up a host name and its aliases.
3268 Previously in this situation something unpredictable would happen;
3269 sometimes it was "internal error: store_reset failed".
3271 18. If a server dropped the connection unexpectedly when an Exim client was
3272 using GnuTLS and trying to read a response, the client delivery process
3273 crashed while trying to generate an error log message.
3275 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3276 string, for example:
3278 warn message = H1: something\nH2: something
3280 the text was added as a single header line from Exim's point of view
3281 though it ended up OK in the delivered message. However, searching for the
3282 second and subsequent header lines using $h_h2: did not work. This has been
3283 fixed. Similarly, if a system filter added multiple headers in this way,
3284 the routers could not see them.
3286 20. Expanded the error message when iplsearch is called with an invalid key to
3287 suggest using net-iplsearch in a host list.
3289 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3290 ACLs are no longer actually imposed (and a message to that effect is
3293 22. If a "gecos" field in a passwd entry contained escaped characters, in
3294 particular, if it contained a \" sequence, Exim got it wrong when building
3295 a From: or a Sender: header from that name. A second bug also caused
3296 incorrect handling when an unquoted " was present following a character
3297 that needed quoting.
3299 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3300 was not being matched caselessly.
3302 24. Arranged for all hyphens in the exim.8 source to be escaped with
3305 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3306 properly. Recipient callout cache records were still being keyed to include
3307 the sender, even when use_sender was set false. This led to far more
3308 callouts that were necessary. The sender is no longer included in the key
3309 when use_sender is false.
3311 26. Added "control = submission" modifier to ACLs.
3313 27. Added the ${base62d: operator to decode base 62 numbers.
3315 28. dnsdb lookups can now access SRV records.
3317 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3318 the configuration file.
3320 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3321 (-v) mode. This makes the output for a verbose queue run more intelligible.
3323 31. Added a use_postmaster feature to recipient callouts.
3325 32. Added the $body_zerocount variable, containing the number of binary zero
3326 bytes in the message body.
3328 33. The time of last modification of the "new" subdirectory is now used as the
3329 "mailbox time last read" when there is a quota error for a maildir
3332 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3334 35. Added +ignore_unknown as a special item in host lists.
3336 36. Code for decoding IPv6 addresses in host lists is now included, even if
3337 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3338 address was recognized as an IP address, but was then not correctly decoded
3339 into binary, causing unexpected and incorrect effects when compared with
3346 1. Very minor rewording of debugging text in manualroute to say "list of
3347 hosts" instead of "hostlist".
3349 2. If verify=header_syntax was set, and a header line with an unqualified
3350 address (no domain) and a large number of spaces between the end of the
3351 name and the colon was received, the reception process suffered a buffer
3352 overflow, and (when I tested it) crashed. This was caused by some obsolete
3353 code that should have been removed. The fix is to remove it!
3355 3. When running in the test harness, delay a bit after writing a bounce
3356 message to get a bit more predictability in the log output.
3358 4. Added a call to search_tidyup() just before forking a reception process. In
3359 theory, someone could use a lookup in the expansion of smtp_accept_max_
3360 per_host which, without the tidyup, could leave open a database connection.
3362 5. Added the variables $recipient_data and $sender_data which get set from a
3363 lookup success in an ACL "recipients" or "senders" condition, or a router
3364 "senders" option, similar to $domain_data and $local_part_data.
3366 6. Moved the writing of debug_print from before to after the "senders" test
3369 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3370 problems for message scanning, either using a data ACL, or using
3371 local_scan() because the Received: header was not generated till after they
3372 were called (in order to set the time as the time of reception completion).
3373 I have revised the way this works. The header is now generated after the
3374 body is received, but before the ACL or local_scan() are called. After they
3375 are run, the timestamp in the header is updated.
3381 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3382 before starting a queue runner without re-exec. This happened only when
3383 deliver_drop_privilege was set or when the Exim user was set to root. The
3384 effect of the bug was that timeouts during subsequent deliveries caused
3385 crashes instead of being properly handled. The handler is now left at its
3386 default (and expected) setting.
3388 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3389 message, again when deliver_drop_privilege is set or Exim is run as root.
3390 The bug described in (1) was not present in this case, but the tidying up
3391 of the other signals was missing. I have made the two cases consistent.
3393 3. The ignore_target_hosts setting on a manualroute router was being ignored
3394 for hosts that were looked up using the /MX notation.
3396 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3399 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3400 operated on the sender address. After changing the $sender_address to <>
3401 for the sender address verify, Exim was re-instated it as the original
3402 (before rewriting) address, but remembering that it had rewritten it, so it
3403 wasn't rewriting it again. This bug also had the effect of breaking the
3404 sender address verification caching when the sender address was rewritten.
3406 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3407 This has been changed so that if the ip literal address matches
3408 ignore_target_hosts, the router declines.
3410 7. Added expansion conditions match_domain, match_address, and match_local_
3411 part (NOT match_host).
3413 8. The placeholder for the Received: header didn't have a length field set.
3415 9. Added code to Exim itself and to exim_lock to test for a specific race
3416 condition that could lead to file corruption when using MBX delivery. The
3417 issue is with the lockfile that is created in /tmp. If this file is removed
3418 after a process has opened it but before that process has acquired a lock,
3419 there is the potential for a second process to recreate the file and also
3420 acquire a lock. This could lead to two Exim processes writing to the file
3421 at the same time. The added code performs the same test as UW imapd; it
3422 checks after acquiring the lock that its file descriptor still refers to
3423 the same named file.
3425 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3426 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3429 11. Added the smtp_active_hostname option. If used, this will typically be made
3430 to depend on the incoming interface address. Because $interface_address is
3431 not set up until the daemon has forked a reception process, error responses
3432 that can happen earlier (such as "too many connections") no longer contain
3435 12. If an expansion in a condition on a "warn" statement fails because a lookup
3436 defers, the "warn" statement is abandoned, and the next ACL statement is
3437 processed. Previously this caused the whole ACL to be aborted.
3439 13. Added the iplsearch lookup type.
3441 14. Added ident_timeout as a log selector.
3443 15. Added tls_certificate_verified as a log selector.
3445 16. Added a global option tls_require_ciphers (compare the smtp transport
3446 option of the same name). This controls incoming TLS connections.
3448 17. I finally figured out how to make tls_require_ciphers do a similar thing
3449 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3450 before starting the TLS session.
3452 18. Tabs are now shown as \t in -bP output.
3454 19. If the log selector return_path_on_delivery was set, Exim crashed when
3455 bouncing a message because it had too many Received: header lines.
3457 20. If two routers both had headers_remove settings, and the first one included
3458 a superfluous trailing colon, the final name in the first list and the
3459 first name in the second list were incorrectly joined into one item (with a
3460 colon in the middle).
3466 1. Added -C and -D options to the exinext utility, mainly to make it easier
3467 to include in the automated testing, but these could be helpful when
3468 multiple configurations are in use.
3470 2. The exinext utility was not formatting the output nicely when there was
3471 an alternate port involved in the retry record key, nor when there was a
3472 message id as well (for retries that were specific to a specific message
3473 and a specific host). It was also confused by IPv6 addresses, because of
3474 the additional colons they contain. I have fixed the IPv4 problem, and
3475 patched it up to do a reasonable job for IPv6.
3477 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3478 delivery, the log line now contains "pipelined" if PIPELINING was used.
3480 4. An SMTP transport process used to panic and die if the bind() call to set
3481 an explicit outgoing interface failed. This has been changed; it is now
3482 treated in the same way as a connect() failure.
3484 5. A reference to $sender_host_name in the part of a conditional expansion
3485 that was being skipped was still causing a DNS lookup. This no longer
3488 6. The def: expansion condition was not recognizing references to header lines
3489 that used bh_ and bheader_.
3491 7. Added the _cache feature to named lists.
3493 8. The code for checking quota_filecount in the appendfile transport was
3494 allowing one more file than it should have been.
3496 9. For compatibility with Sendmail, the command line option
3504 and sets the incoming protocol and host name (for trusted callers). The
3505 host name and its colon can be omitted when only the protocol is to be set.
3506 Note the Exim already has two private options, -pd and -ps, that refer to
3507 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3508 "s", but I don't think that's a major issue.
3510 10. A number of refactoring changes to the code, none of which should affect
3513 (a) The number of logging options was getting close to filling up the
3514 32-bit word that was used as a bit map. I have split them into two classes:
3515 those that are passed in the argument to log_write(), and those that are
3516 only ever tested independently outside of that function. These are now in
3517 separate 32-bit words, so there is plenty of room for expansion again.
3518 There is no change in the user interface or the logging behaviour.
3520 (b) When building, for example, log lines, the code previously used a
3521 macro that called string_cat() twice, in order to add two strings. This is
3522 not really sufficiently general. Furthermore, there was one instance where
3523 it was actually wrong because one of the argument was used twice, and in
3524 one call a function was used. (As it happened, calling the function twice
3525 did not affect the overall behaviour.) The macro has been replaced by a
3526 function that can join an arbitrary number of extra strings onto a growing
3529 (c) The code for expansion conditions now uses a table and a binary chop
3530 instead of a serial search (which was left over from when there were very
3531 few conditions). Also, it now recognizes conditions like "pam" even when
3532 the relevant support is not compiled in: a suitably worded error message is
3533 given if an attempt is made to use such a condition.
3535 11. Added ${time_interval:xxxxx}.
3537 12. A bug was causing one of the ddress fields not to be passed back correctly
3538 from remote delivery subprocesses. The field in question was not being
3539 subsequently used, so this caused to problems in practice.
3541 13. Added new log selectors queue_time and deliver_time.
3543 14. Might have fixed a bug in maildirsizefile handling that threw up
3544 "unexpected character" debug warnings, and recalculated the data
3545 unnecessarily. In any case, I expanded the warning message to give more
3548 15. Added the message "Restricted characters in address" to the statements in
3549 the default ACL that block characters like @ and % in local parts.
3551 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3552 Three changes have been made:
3554 (a) There was a serious bug; a negative response to MAIL caused the whole
3555 recipient domain to be cached as invalid, thereby blocking all messages
3556 to all local parts at the same domain, from all senders. This bug has
3557 been fixed. The domain is no longer cached after a negative response to
3558 MAIL if the sender used is not empty.
3560 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3563 (c) A new callout option, "use_sender" has been added for people who want
3564 the modified behaviour.
3570 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3573 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3574 indeed breaks things for older releases.
3576 3. Added additional logging to the case where there is a problem reading data
3577 from a filter that is running in a subprocess using a pipe, in order to
3578 try to track down a specific problem.
3580 4. Testing facility fudge: when running in the test harness and attempting
3581 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3582 getting "No route to host". Convert this to a timeout.
3584 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3587 6. Some OS don't have socklen_t but use size_t instead. This affects the
3588 fifth argument of getsockopt() amongst other things. This is now
3589 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3590 can be set for individual OS. I have set it for SunOS5, OSF1, and
3591 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3592 some earlier ones do not.
3594 7. Change 4.30/15 was not doing the test caselessly.
3596 8. The standard form for an IPv6 address literal was being rejected by address
3597 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3598 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3599 this, as well as the form without the "IPv6" on the front (but only when
3600 address literals are enabled, of course).
3602 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3604 10. Exim crashed if a message with an empty sender address specified by -f
3605 encountered a router with an errors_to setting. This could be provoked only
3606 by a command such as
3610 where an empty string was supplied; "<>" did not hit this bug.
3612 11. Installed PCRE release 4.5.
3614 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3615 remained set. It is now erased.
3617 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3618 times from message ids (which are base 36 rather than the normal 62).
3620 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3621 were being counted as actual protocol errors, and logged if the log
3622 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3623 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3624 and DATA following a set of rejected RCPTs do not count as protocol errors.
3625 In other words, Exim assumes they were pipelined, though this may not
3626 actually be the case. Of course, in all cases the client gets an
3627 appropriate error code.
3629 15. If a lookup fails in an ACL condition, a message about the failure may
3630 be available; it is used if testing the ACL cannot continue, because most
3631 such messages specify what the cause of the deferral is. However, some
3632 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3633 that caused an old message to be retained and used if a later statement
3634 caused a defer, replacing the real cause of the deferral.
3636 16. If an IP address had so many PTR records that the DNS lookup buffer
3637 was not large enough to hold them, Exim could crash while trying to process
3638 the truncated data. It now detects and logs this case.
3640 17. Further to 4.21/58, another change has been made: if (and only if) the
3641 first line of a message (the first header line) ends with CRLF, a bare LF
3642 in a subsequent header line has a space inserted after it, so as not to
3643 terminate the header.
3645 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3646 unnecessarily, used atoi() instead of strtol(), and didn't check the
3647 termination when getting file sizes from file names by regex.
3649 19. Completely re-implemented the support for maildirsize files, in the light
3650 of a number of problems with the previous contributed implementation
3651 (4.30/29). In particular:
3653 . If the quota is zero, the maildirsize file is maintained, but no quota is
3656 . If the maildir directory does not exist, it is created before any attempt
3657 to write a maildirsize file.
3659 . The quota value in the file is just a cache; if the quota is changed in
3660 the transport, the new value overrides.
3662 . A regular expression is available for excluding directories from the
3665 20. The autoreply transport checks the characters in options that define the
3666 message's headers; it allows continued headers, but it was checking with
3667 isspace() after an embedded newline instead of explicitly looking for a
3670 21. If all the "regular" hosts to which an address was routed had passed their
3671 expiry times, and had not reached their retry times, the address was
3672 bounced, even if fallback hosts were defined. Now Exim should go on to try
3675 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3676 equivalent code in the SMTP transport. Some hosts send humungous responses
3677 to HELO/EHLO, more than 1024 it seems.
3679 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3680 gives compiler warnings in some environments. I've now done it "properly",
3683 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3684 (because of problems with the built-in one) was declared to return uschar *
3685 instead of char *, causing compiler failure.
3687 25. Fixed a file descriptor leak when processing alias/forward files.
3689 26. Fixed a minor format string issue in dbfn.c.
3691 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3693 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3694 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3697 29. When a sender address is verified, it is cached, to save repeating the test
3698 when there is more than one recipient in a message. However, when the
3699 verification involves a callout, it is possible for different callout
3700 options to be set for different recipients. It is too complicated to keep
3701 track of this in the cache, so now Exim always runs a verification when a
3702 callout is required, relying on the callout cache for the optimization.
3703 The overhead is duplication of the address routing, but this should not be
3706 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3707 to be verified with callout=postmaster, and the main callout worked but the
3708 postmaster check failed, the verification correctly failed. However, if a
3709 subsequent RCPT command asked for sender verification *without* the
3710 postmaster check, incorrect caching caused this verification also to fail,
3713 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3714 it was not caching the DNS options (qualify_single, search_parents) that
3715 were used when the lookup failed. A subsequent lookup with different
3716 options therefore always gave the same answer, though there were cases
3717 where it should not have. (Example: a "domains = !$mx_any" option on a
3718 dnslookup router: the "domains" option is always processed without any
3719 widening, but the router might have qualify_single set.) Now Exim uses the
3720 cached value only when the same options are set.
3722 32. Added John Jetmore's "exipick" utility to the distribution.
3724 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3725 than a timeout (e.g. a certificate is required, and is not provided), an
3726 Exim server now closes the connection immediately. Previously it waited for
3727 the client to close - but if the client is SSL, it seems that they each
3728 wait for each other, leading to a delay before one of them times out.
3730 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3731 maintained 0.8.x compatibility because I don't think many are using it, and
3732 it is clearly obsolete.
3734 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3737 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3738 client certificate was expired. A simple patch fixes this, though I don't
3739 understand the full logic of why the verify callback is called multiple
3742 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3743 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3744 which causes problems with some clients (such as the Certicom SSL Plus
3745 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3746 disables the coutermeasure allowing Eudora to connect."
3748 38. Exim was not checking that a write() to a log file succeeded. This could
3749 lead to Bad Things if a log got too big, in particular if it hit a file
3750 size limit. Exim now panics and dies if it cannot write to a log file, just
3751 as it does if it cannot open a log file.
3753 39. Modified OS/Makefile-Linux so that it now contains
3755 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3757 The two -D definitions ensure that Exim is compiled with large file
3758 support, which makes it possible to handle log files that are bigger than
3761 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3762 instance) a domain was checked against a named list that involved a lookup,
3763 causing $domain_data to be set, then another domain was checked against the
3764 same list, then the first domain was re-checked, the value of $domain_data
3765 after the final check could be wrong. In particular, if the second check
3766 failed, it could be set empty. This bug probably also applied to
3769 41. The strip_trailing_dot option was not being applied to the address given
3770 with the -f command-line option.
3772 42. The code for reading a message's header from the spool was incrementing
3773 $received_count, but never initializing it. This meant that the value was
3774 incorrect (doubled) while delivering a message in the same process in which
3775 it was received. In the most common configuration of Exim, this never
3776 happens - a fresh exec is done - but it can happen when
3777 deliver_drop_privilege is set.
3779 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3780 it now includes up to 150 characters of the unexpected data in the log
3783 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3784 and building data strings. The size of both of these buffers was 10 000
3785 bytes - far larger than anybody would *ever* want, thought I. Needless to
3786 say, somebody hit the limit. I have increased the maximum line length to
3787 20 000 and the maximum data length of concatenated lines to 100 000. I have
3788 also fixed two bugs, because there was no checking on these buffers. Tsk,
3789 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3790 buffer is too small.
3792 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3793 lsearch lookups. Now it does.
3795 46. When parsing a route_list item in a manualroute router, a fixed-length
3796 buffer was used for the list of hosts. I made this 1024 bytes long,
3797 thinking that nobody would ever have a list of hosts that long. Wrong.
3798 Somebody had a whole pile of complicated expansion conditions, and the
3799 string was silently truncated, leading to an expansion error. It turns out
3800 that it is easier to change to an unlimited length (owing to other changes
3801 that have happened since this code was originally written) than to build
3802 structure for giving a limitation error. The length of the item that
3803 expands into the list of hosts is now unlimited.
3805 47. The lsearch lookup could not handle data where the length of text line was
3806 more than 4095 characters. Such lines were truncated, leading to shortened
3807 data being returned. It should now handle lines of any length.
3809 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3810 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3813 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3814 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3815 no longer happens in comment lines. A list of replacements is now placed
3816 at the head of all of the source files, except those whose only change is
3817 to replace PERL_COMMAND in the very first #! line.
3819 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3820 messages on the queue, with a bottom-up merge sort, using code contributed
3821 by Michael Haardt. This should make operations like -bp somewhat faster on
3822 large queues. It won't affect queue runners, except when queue_run_in_order
3825 51. Installed eximstats 1.31 in the distribution.
3827 52. Added support for SRV lookups to the dnslookup router.
3829 53. If an ACL referred to $message_body or $message_body_end, the value was not
3830 reset for any messages that followed in the same SMTP session.
3832 54. The store-handling optimization for building very long strings was not
3833 differentiating between the different store pools. I don't think this
3834 actually made any difference in practice, but I've tidied it.
3836 55. While running the routers to verify a sender address, $sender_address
3837 was still set to the sender address. This is wrong, because when routing to
3838 send a bounce to the sender, it would be empty. Therefore, I have changed
3839 it so that, while verifying a sender address, $sender_address is set to <>.
3840 (There is no change to what happens when verifying a recipient address.)
3842 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3843 target A or AAAA records (if not already returned) without resetting the
3844 qualify_single or search_parents options of the DNS resolver. These are
3845 inappropriate in this case because the targets of MX and SRV records must
3846 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3847 target that, when qualified, matched something in the local domain. These
3848 two options are now turned off when doing these lookups.
3850 57. It seems that at least some releases of Reiserfs (which does not have the
3851 concept of a fixed number of inodes) returns zero and not -1 for the
3852 number of available inodes. This interacted badly with check_spool_inodes,
3853 which assumed that -1 was the "no such thing" setting. What I have done is
3854 to check that the total number of inodes is greater than zero before doing
3855 the test of how many are available.
3857 58. When a "warn" ACL statement has a log_message modifier, the message is
3858 remembered, and not repeated. This is to avoid a lot of repetition when a
3859 message has many recipients that cause the same warning to be written.
3860 Howewer, Exim was preserving the list of already written lines for an
3861 entire SMTP session, which doesn't seem right. The memory is now reset if a
3862 new message is started.
3864 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3865 debugging output unless log_rewrite was also set.
3867 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3868 of (int)(handle) when we know that handle contains (void *)(-1).
3870 61. The Exim daemon panic-logs an error return when it closes the incoming
3871 connection. However "connection reset by peer" seems to be common, and
3872 isn't really an error worthy of noting specially, so that particular error
3875 62. When Exim is trying to find all the local interfaces, it used to panic and
3876 die if the ioctl to get the interface flags failed. However, it seems that
3877 on at least one OS (Solaris 9) it is possible to have an interface that is
3878 included in the list of interfaces, but for which you get a failure error
3879 for this call. This happens when the interface is not "plumbed" into a
3880 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3881 failure of the "get flags" call assumes that the interface is down.
3883 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3884 makes life easier for people who are doing arithmetic on fields extracted
3885 from dates, where you often get leading zeros that should not be
3886 interpreted as octal.
3888 64. Added qualify_domain to the redirect router, to override the global
3891 65. If a pathologically long header line contained very many addresses (the
3892 report of this problem mentioned 10 000) and each of them was rewritten,
3893 Exim could use up a very large amount of memory. (It kept on making new
3894 copies of the header line as it rewrote, and never released the old ones.)
3895 At the expense of a bit more processing, the header rewriting function has
3896 been changed so that it no longer eats memory in this way.
3898 66. The generation of the Received: header has been moved from the time that a
3899 message starts to be received, to the time that it finishes. The timestamp
3900 in the Received: header should now be very close to that of the <= log
3901 line. There are two side-effects of this change:
3903 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3904 logged header lines no longer include the local Received: line, because
3905 it has not yet been created. The same applies to a copy of the message
3906 that is returned to a non-SMTP sender when a message is rejected.
3908 (b) When a filter file is tested using -bf, no additional Received: header
3909 is added to the test message. After some thought, I decided that this
3912 This change does not affect the value of $received_for. It is still set
3913 after address rewriting, but before local_scan() is called.
3915 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3917 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3918 gave an unhelpful panic error message, and a defer error. I have managed to
3919 change this behaviour so that it now rejects any supplied certificate,
3920 which seems right, as the list of acceptable certificates is empty.
3922 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3923 gave an unhelpful defer error. I have not managed to make this reject any
3924 supplied certificates, but the error message it gives is "no certificate
3925 supplied", which is not helpful.
3927 70. exigrep's output now also includes lines that are not associated with any
3928 message, but which match the given pattern. Implemented by a patch from
3929 Martin Sluka, which also tidied up the Perl a bit.
3931 71. Recipient callout verification, like sender verification, was using <> in
3932 the MAIL FROM command. This isn't really the right thing, since the actual
3933 sender may affect whether the remote host accepts the recipient or not. I
3934 have changed it to use the actual sender in the callout; this means that
3935 the cache record is now keyed on a recipient/sender pair, not just the
3936 recipient address. There doesn't seem to be a real danger of callout loops,
3937 since a callout by the remote host to check the sender would use <>.
3938 [SEE ABOVE: changed after hitting problems.]
3940 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3941 temporary errors. However, in the case of such a code being given after
3942 the end of a data transmission (i.e. after ".") Exim was failing to write
3943 a retry record for the message. (Yes, there was some broken host that was
3944 actually sending 8xx at this point.)
3946 73. An unknown lookup type in a host list could cause Exim to panic-die when
3947 the list was checked. (An example that provoked this was putting <; in the
3948 middle of a list instead of at the start.) If this happened during a DATA
3949 ACL check, a -D file could be left lying around. This kind of configuration
3950 error no longer causes Exim to die; instead it causes a defer errror. The
3951 incident is still logged to the main and panic logs.
3953 74. Buglet left over from Exim 3 conversion. The message "too many messages
3954 in one connection" was written to the rejectlog but not the mainlog, except
3955 when address rewriting (yes!) was being logged.
3957 75. Added write_rejectlog option.
3959 76. When a system filter was run not as root (that is, when system_filter_user
3960 was set), the values of the $n variables were not being returned to the
3961 main process; thus, they were not subsequently available in the $sn
3964 77. Added +return_path_on_delivery log selector.
3966 78. A connection timeout was being treated differently from recipients deferred
3967 when testing hosts_max_try with a message that was older than the host's
3968 retry timeout. (The host should not be counted, thus allowing all hosts to
3969 be tried at least once before bouncing.) This may have been the cause of an
3970 occasionally reported bug whereby a message would remain on the queue
3971 longer than the retry timeout, but would be bounced if a delivery was
3972 forced. I say "may" because I never totally pinned down the problem;
3973 setting up timeout/retry tests is difficult. See also the next item.
3975 79. The ultimate address timeout was not being applied to errors that involved
3976 a combination of host plus message (for example, a timeout on a MAIL
3977 command). When an address resolved to a number of possible hosts, and they
3978 were not all tried for each delivery (e.g. because of hosts_max_try), a
3979 message could remain on the queue longer than the retry timeout.
3981 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3984 81. Fixed an obscure SMTP outgoing bug which required at least the following
3985 conditions: (a) there was another message waiting for the same server;
3986 (b) the server returned 5xx to all RCPT commands in the first message so
3987 that the message was not completed; (c) the server dropped the connection
3988 or gave a negative response to the RSET that Exim sends to abort the
3989 transaction. The observed case was a dropped connection after DATA that had
3990 been sent in pipelining mode. That is, the server had advertised PIPELINING
3991 but was not implementing it correctly. The effect of the bug was incorrect
3992 behaviour, such as trying another host, and this could lead to a crash.
3998 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3999 and daemon.c were passed as pointers to ints; they should have been
4000 pointers to socklen_t variables (which are typically unsigned ints).
4002 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4005 3. Fixed a really odd bug that affected only the testing scheme; patching a
4006 certain fixed string in the binary changed the value of another string that
4007 happened to be identical to the end of the original first string.
4009 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4010 name", it returns that address as the IP address. On some operating
4011 systems (e.g. Solaris), it also passes back the IP address string as the
4012 "host name". However, on others (e.g. Linux), it passes back an empty
4013 string. Exim wasn't checking for this, and was changing the host name to an
4014 empty string, assuming it had been canonicized.
4016 5. Although rare, it is permitted to have more than one PTR record for a given
4017 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4018 all the names associated with an address, because they do in Solaris.
4019 However, it seems that they do not in Linux for data that comes from the
4020 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4021 I found this out when I moved to a new Linux workstation and tried to run
4022 the Exim test suite.
4024 To get round this problem I have changed the code so that it now does its
4025 own call to the DNS to look up PTR records when searching for a host name.
4026 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4027 addresses that are only in /etc/hosts are still found.
4029 This behaviour is, however, controlled by an option called host_lookup_
4030 order, which defaults to "bydns:byaddr". If people want to use the other
4031 order, or indeed, just use one or the other means of lookup, they can
4032 specify it in this variable.
4034 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4035 some operating systems, this comes back from gethostbyaddr() as an empty
4036 string, and this is what Exim used to test for. However, it seems that in
4037 other systems, "." is yielded. Exim now tests for this case too.
4039 7. The values of check_spool_space and check_log_space are now held internally
4040 as a number of kilobytes instead of an absolute number of bytes. If a
4041 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4042 kilobyte. This means that much larger values can be stored.
4044 8. Exim monitor: an attempt to get the action menu when not actually pointing
4045 at a message produces an empty menu entitled "No message selected". This
4046 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4047 no entries in it ("Shell widget menu has zero width and/or height"). So I
4048 have added a single, blank menu entry in this case.
4050 9. Added ${quote_local_part.
4052 10. MIME decoding is now applied to the contents of Subject: header lines when
4055 11. Now that a reference to $sender_host_address automatically causes a reverse
4056 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4057 host lookup before query-style lookups in lists that might use this
4058 variable. This has therefore been abolished, and the "net-" prefix is no
4059 longer necessary for query-style lookups.
4061 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4062 have been a typo for LFLAGS, so it has been changed.
4064 13. The install script calls Exim with "-C /dev/null" in order to find the
4065 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4066 to be output. Howeve, since Exim outputs its version number before the
4067 error, it didn't break the script. It just looked ugly. I fixed this by
4068 always allowing "-C /dev/null" if the caller is root.
4070 14. Ignore overlarge ACL variable number when reading spool file - insurance
4071 against a later release with more variables having written the file.
4073 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4074 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4075 this, as well as the form without the "IPv6" on the front.
4077 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4078 OS/Makefile-Darwin file.
4080 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4081 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4083 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4084 variable. However, some OS use ints for inodes. I've added cast to long int
4085 to get rid of the compiler warning.
4087 19. I had forgotten to lock out "/../" in configuration file names when
4088 ALT_CONFIG_PREFIX was set.
4090 20. Routers used for verification do not need to specify transports. However,
4091 if such a router generated a host list, and callout was configured, Exim
4092 crashed, because it could not find a port number from the (non-existent)
4093 transport. It now assumes port 25 in this circumstance.
4095 21. Added the -t option to exigrep.
4097 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4098 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4099 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4100 them is set, LOOKUP_LSEARCH is forced.
4102 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4104 24. Added sender and host information to the "rejected by local_scan()" log
4105 line; previously there was no indication of these.
4107 25. Added .include_if_exists.
4109 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4110 Linux. It turns out that not all file systems support this. Apparently some
4111 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4112 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4113 not supported on the file descriptor, is now ignored when Exim is trying to
4114 sync a directory. This applies only to Linux.
4116 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4118 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4119 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4122 29. Added support for maildirsize files from supplied patch (modified a bit).
4124 30. The use of :fail: followed by an empty string could lead Exim to respond to
4125 sender verification failures with (e.g.):
4127 550 Verification failed for <xxx>
4128 550 Sender verify failed
4130 where the first response line was missing the '-' that indicates it is not
4131 the final line of the response.
4133 31. The loop for finding the name of the user that called Exim had a hardwired
4134 limit of 10; it now uses the value of finduser_retries, which is used for
4135 all other user lookups.
4137 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4140 33. Exim was neglecting to zero errno before one call of strtol() when
4141 expanding a string and expecting an integer value. On some systems this
4142 resulted in spurious "integer overflow" errors. Also, it was casting the
4143 result into an int without checking.
4145 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4146 did not work. The code looks as if it has *never* worked, though it appears
4147 to have been documented since at least releast 1.62. I have made it work.
4149 35. The "timeout_DNS" error in retry rules, also documented since at least
4150 1.62, also never worked. As it isn't clear exactly what this means, and
4151 clearly it isn't a major issue, I have abolished the feature by treating it
4152 as "timeout", and writing a warning to the main and panic logs.
4154 36. The display of retry rules for -brt wasn't always showing the error code
4157 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4158 timeout_connect_A, timeout_connect_MX.
4160 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4161 to the empty sender.
4163 39. The daemon was not analysing the content of -oX till after it had closed
4164 stderr and disconnected from the controlling terminal. This meant that any
4165 syntax errors were only noted on the panic log, and the return code from
4166 the command was 0. By re-arranging the code a little, I've made the
4167 decoding happen first, so such errors now appear on stderr, and the return
4168 code is 1. However, the actual setting up of the sockets still happens in
4169 the disconnected process, so errors there are still only recorded on the
4172 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4173 connections (as happens on some IP stacks) was logged at start up time as
4174 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4175 it from "IPv6 and IPv4", which means that two separate sockets are being
4178 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4179 says whether AF_INET or AF_INET6 was passed as an argument.
4181 42. Exiwhat output was messed up when time zones were included in log
4184 43. Exiwhat now gives more information about the daemon's listening ports,
4185 and whether -tls-on-connect was used.
4187 44. The "port" option of the smtp transport is now expanded.
4189 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4190 silently ignored. Now an error message is written to the main and panic
4193 46. There's a new ACL modifier called "logwrite" which writes to a log file
4194 as soon as it is encountered.
4196 47. Added $local_user_uid and $local_user_gid at routing time.
4198 48. Exim crashed when trying to verify a sender address that was being
4201 49. Exim was recognizing only a space character after ".include". It now also
4202 recognizes a tab character.
4204 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4205 extracting the relevant information from the specification. The man page no
4206 longer contains scrambled data for the -d option, and I've added a section
4207 at the front about calling Exim under different names.
4209 51. Added "extra_headers" argument to the "mail" command in filter files.
4211 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4214 53. Installed eximstats 1.29.
4216 54. Added transport_filter_timeout as a generic transport option.
4218 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4219 Cc: header lines. This was required by RFC 822, but it not required by RFC
4222 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4223 incoming messages that did not have them. Now it does so only if the
4224 message originates locally, that is, if there is no associated remote host
4225 address. When Resent- header lines are present, this applies to the Resent-
4226 lines rather than the non-Resent- lines.
4228 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4229 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4231 58. Messages for configuration errors now include the name of the main
4232 configuration file - useful now that there may be more than one file in a
4233 list (.included file names were always shown).
4235 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4236 for those rare installations that do not start the daemon as root or run it
4237 setuid root. I've cut out the call to initgroups() if the daemon is not
4240 60. The Exim user and group can now be bound into the binary as text strings
4241 that are looked up at the start of Exim's processing.
4243 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4245 62. Added $mailstore_basename variable.
4247 63. Installed patch to sieve.c from Michael Haardt.
4249 64. When Exim failed to open the panic log after failing to open the main log,
4250 the original message it was trying to log was written to stderr and debug
4251 output, but if they were not available (the usual case in production), it
4252 was lost. Now it is written to syslog before the two lines that record the
4253 failures to open the logs.
4255 65. Users' Exim filters run in subprocesses under the user's uid. It is
4256 possible for a "deliver" command or an alias in a "personal" command to
4257 provoke an address rewrite. If logging of address rewriting is configured,
4258 this fails because the process is not running as root or exim. There may be
4259 a better way of dealing with this, but for the moment (because 4.30 needs
4260 to be released), I have disabled address rewrite logging when running a
4261 filter in a non-root, non-exim process.
4267 1. The buildconfig auxiliary program wasn't quoting the value set for
4268 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4269 not defined. This bug was masked in 4.22 by the effect that was fixed in
4272 2. Some messages that were rejected after a message id was allocated were
4273 shown as "incomplete" by exigrep. It no longer does this for messages that
4274 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4276 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4277 have allow_domain_literals set, the ID did not get logged in the <= line.
4278 Domain literals are now always recognized in Message-ID: header lines.
4280 4. The first argument for a ${extract expansion item is the key name or field
4281 number. Leading and trailing spaces in this item were not being ignored,
4282 causing some misleading effects.
4284 5. When deliver_drop_privilege was set, single queue runner processes started
4285 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4286 same command in the process it spins off) were not dropping privilege.
4288 6. When the daemon running as "exim" started a queue runner, it always
4289 re-executed Exim in the spun-off process. This is a waste of effort when
4290 deliver_drop_privilege is set. The new process now just calls the
4291 queue-runner function directly.
4297 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4300 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4301 ignored. Though the use of -oP was forcing the writing of a pid file, it
4302 was always written to the default place.
4304 3. If the message "no IP address found for host xxxx" is generated during
4305 incoming verification, it is now followed by identification of the incoming
4306 connection (so you can more easily find what provoked it).
4308 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4310 5. Added some features to "harden" Exim a bit more against certain attacks:
4312 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4313 be put in Local/Makefile. This is like the never_users runtime option,
4314 but it cannot be overridden. The default setting is "root".
4316 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4317 prefix string with which any file named in a -C command line option
4320 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4321 is retained for -C and -D only if the caller of Exim is root. Without
4322 it, the exim user may also use -C and -D and retain privilege.
4324 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4325 command line option is disabled.
4327 6. Macro names set by the -D option must start with an upper case letter, just
4328 like macro names defined in the configuration file.
4330 7. Added "dereference=" facility to LDAP.
4332 8. Two instances of the typo "uknown" in the source files are fixed.
4334 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4335 the Configure-Makefile script screwed up while processing it.
4337 10. Incorporated PCRE 4.4.
4339 11. The SMTP synchronization check was not operating right at the start of an
4340 SMTP session. For example, it could not catch a HELO sent before the client
4341 waited for the greeting. There is now a check for outstanding input at the
4342 point when the greeting is written. Because of the duplex, asynchronous
4343 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4344 way, but not yet received, when the check is performed.
4346 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4347 on TCP/IP sockets, because this apparently causes some broken clients to
4350 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4351 unchanged) from the Cygwin maintainer.
4353 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4354 was supported instead of testing for mbx. Effectively a typo.
4356 15. The spa authenticator server code was not checking that the input it
4357 received was valid base64.
4359 16. The debug output line for the "set" modifier in ACLs was not showing the
4360 name of the variable that was being set.
4362 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4364 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4365 lookup on its own. Something else was needed to trigger the lookup. For
4366 example, a match in host_lookup or the need for a host name in a host list.
4367 Now, if $sender_host_name is referenced and the host name has not yet been
4368 looked up, a lookup is performed. If the lookup fails, the variable remains
4369 empty, and $host_lookup_failed is set to "1".
4371 19. Added "eqi" as a case-independent comparison operator.
4373 20. The saslauthd authentication condition could segfault if neither service
4374 nor realm was specified.
4376 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4377 error message that was logged was misleading, and incoming SMTP
4378 connections were dropped. The message is now more accurate, and temporary
4379 errors are given to SMTP connections.
4381 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4382 (except RSET) with a 421 error, until QUIT is received. However, it was
4383 failing to send a response to QUIT.
4385 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4386 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4387 ACL is now run after the helo_try_verify_hosts code.
4389 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4390 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4391 case-independent, so other case variants are also recognized). Apparently
4392 some systems use these upper case variants.
4394 25. If more than two messages were waiting for the same host, and a transport
4395 filter was specified for the transport, Exim sent two messages over the
4396 same TCP/IP connection, and then failed with "socket operation on non-
4397 socket" when it tried to send the third.
4399 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4401 27. The extern definition of crypt16() in expand.c was not being excluded when
4402 the OS had its own crypt16() function.
4404 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4405 as a preferred synonym for return_size_limit, both as an option and as an
4408 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4410 30. Changed the default configuration ACL to relax the local part checking rule
4411 for addresses that are not in any local domains. For these addresses,
4412 slashes and pipe symbols are allowed within local parts, but the sequence
4413 /../ is explicitly forbidden.
4415 31. SPA server authentication was not clearing the challenge buffer before
4418 32. log_message in a "warn" ACL statement was writing to the reject log as
4419 well as to the main log, which contradicts the documentation and doesn't
4420 seem right (because no rejection is happening). So I have stopped it.
4422 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4423 However, I am unable to do any testing of this.
4425 34. Fixed an infelicity in the appendfile transport. When checking directories
4426 for a mailbox, to see if any needed to be created, it was accidentally
4427 using path names with one or more superfluous leading slashes; tracing
4428 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4430 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4431 contents are added to the log line that is written for every discarded
4432 recipient. (Previously a log_message setting was ignored.)
4434 36. The ${quote: operator now quotes the string if it is empty.
4436 37. The install script runs exim in order to find its version number. If for
4437 some reason other than non-existence or emptiness, which it checks, it
4438 could not run './exim', it was installing it with an empty version number,
4439 i.e. as "exim-". This error state is now caught, and the installation is
4442 38. An argument was missing from the function that creates an error message
4443 when Exim fails to connect to the socket for saslauthd authentication.
4444 This could cause Exim to crash, or give a corrupted message.
4446 39. Added isip, isip4, and isip6 to ${if conditions.
4448 40. The ACL variables $acl_xx are now saved with the message, and can be
4449 accessed later in routers, transports, and filters.
4451 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4452 strings in the file are not string-expanded.
4454 42. If a MAIL command specified a SIZE value that was too large to fit into an
4455 int variable, the check against message_size_limit failed. Such values are
4456 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4457 day this will have to be increased, but I don't think I want to be around
4458 when emails are that large.
4465 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4466 iconv() is not standard in FreeBSD.
4468 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4469 IPv6 enabled. The observed symptom was a segmentation fault on return from
4470 the function os_common_find_running_interfaces() in src/os.c.
4472 3. In the check_special_case() function in daemon.c I had used "errno" as an
4473 argument name, which causes warnings on some systems. This was basically a
4474 typo, since it was named "eno" in the comments!
4476 4. The code that waits for the clock to tick (at a resolution of some fraction
4477 of a second) so as to ensure message-id uniqueness was always waiting for
4478 at least one whole tick, when it could have waited for less. [This is
4479 almost certainly not relevant at current processor speeds, where it is
4480 unlikely to ever wait at all. But we try to future-proof.]
4482 5. The function that sleeps for a time interval that includes fractions of a
4483 second contained a race. It did not block SIGALRM between setting the
4484 timer, and suspending (a couple of lines later). If the interval was short
4485 and the sigsuspend() was delayed until after it had expired, the suspension
4486 never ended. On busy systems this could lead to processes getting stuck for
4489 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4490 process, before it forks any delivery processes. The open lookup caching
4491 mechanism meant that the open file or database connection was passed into
4492 the delivery process. The problem was that delivery processes always tidy
4493 up cached lookup data. This could cause a problem for the next delivery
4494 process started by the queue runner, because the external queue runner
4495 process does not know about the closure. So the next delivery process
4496 still has data in the lookup cache. In the case of a file lookup, there was
4497 no problem because closing a file descriptor in a subprocess doesn't affect
4498 the parent. However, if the lookup was caching a connection to a database,
4499 the connection was closed, and the second delivery process was likely to
4500 see errors such as "PGSQL: query failed: server closed the connection
4501 unexpectedly". The problem has been fixed by closing all cached lookups
4502 in a queue runner before running a delivery process.
4504 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4505 seem to have the "const" qualifier which it has on other OS. I've
4508 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4509 *of the same type* (client or server) with the same public name that an
4510 error should be diagnosed.
4512 9. When Exim looked up a host name for an IP address, but failed to find the
4513 original IP address when looking up the host name (a safety check), it
4514 output the message "<ip address> does not match any IP for NULL", which was
4515 confusing, to say the least. The bug was that the host name should have
4516 appeared instead of "NULL".
4518 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4519 user that is built into the binary, and the -C or -D options is used, root
4520 privilege is dropped before the configuration file is read. In addition,
4521 logging is switched to stderr instead of the normal log files. If the
4522 configuration then re-defines the Exim user, the unprivileged environment
4523 is probably not what is expected, so Exim logs a panic warning message (but
4526 However, if deliver_drop_privilege is set, the unprivileged state may well
4527 be exactly what is intended, so the warning has been cut out in that case,
4528 and Exim is allowed to try to write to its normal log files.
4534 1. smtp_return_error_details was not giving details for temporary sender
4535 or receiver verification errors.
4537 2. Diagnose a configuration error if two authenticators have the same public
4540 3. Exim used not to create the message log file for a message until the first
4541 delivery attempt. This could be confusing when incoming messages were held
4542 for policy or load reasons. The message log file is now created at the time
4543 the message is received, and an initial "Received" line is written to it.
4545 4. The automatically generated man page for command line options had a minor
4546 bug that caused no ill effects; however, a more serious problem was that
4547 the procedure for building the man page automatically didn't always
4548 operate. Consequently, release 4.20 contains an out-of-date version. This
4549 shouldn't happen again.
4551 5. When building Exim with embedded Perl support, the script that builds the
4552 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4553 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4555 6. The freeze_tell option was not being used for messages that were frozen on
4556 arrival, either by an ACL or by local_scan().
4558 7. Added the smtp_incomplete_transaction log selector.
4560 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4561 was accepting AUTH without a new EHLO.
4563 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4564 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4567 10. Logging of TCP/IP connections (when configured) now happens in the main
4568 daemon process instead of the child process, so that the TCP/IP connection
4569 count is more accurate (but it can never be perfect).
4571 11. The use of "drop" in a nested ACL was not being handled correctly in the
4572 outer ACL. Now, if condition failure induced by the nested "drop" causes
4573 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4574 or "require"), the connection is dropped.
4576 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4577 that yield "discard" can now be used with an "accept" or a "discard" verb,
4578 but an error is generated for any others (because I can't see a useful way
4579 to define what should happen).
4581 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4582 are now processed in the same way as lines in the Exim configuration file.
4583 In particular, continuation lines are supported.
4585 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4587 15. Added -ti meaning -t -i.
4589 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4590 domains, and warn by logging if others are found.
4592 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4593 returned from GIFCONF when Exim is trying to find the list of interfaces on
4594 a host. The code in os.c has been modified to copy each ifreq to an aligned
4595 structure in all cases.
4597 Also, in some cases, the returned ifreq's were being copied to a 'struct
4598 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4599 means the last couple of bytes of an IPv6 address could be chopped if the
4600 ifreq contained only a normal sockaddr (14 bytes storage).
4602 18. Named domain lists were not supported in the hosts_treat_as_local option.
4603 An entry such as +xxxx was not recognized, and was treated as a literal
4606 19. Ensure that header lines added by a DATA ACL are included in the reject log
4607 if the ACL subsequently rejects the message.
4609 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4610 MD5 (which is deprecated).
4612 21. When testing a filter file using -bf, Exim was writing a message when it
4613 took the sender from a "From " line in the message, but it was not doing so
4614 when it took $return_path from a Return-Path: header line. It now does.
4616 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4617 with a valid header line field name (a series of printing characters
4618 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4620 23. Changed "disc" in the source to "disk" to conform to the documentation and
4621 the book and for uniformity.
4623 24. Ignore Sendmail's -Ooption=value command line item.
4625 25. When execve() failed while trying to run a command in a pipe transport,
4626 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4627 could be confused with a return value of 69 from the command itself. This
4628 has been changed to 127, the value the shell returns if it is asked to run
4629 a non-existent command. The wording for the related log line suggests a
4630 non-existent command as the problem.
4632 26. If received_header_text expands to an empty string, do not add a Received:
4633 header line to the message. (Well, it adds a token one on the spool, but
4634 marks it "old" so that it doesn't get used or transmitted.)
4636 27. Installed eximstats 1.28 (addition of -nt option).
4638 28. There was no check for failure on the call to getsockname() in the daemon
4639 code. This can fail if there is a shortage of resources on the system, with
4640 ENOMEM, for example. A temporary error is now given on failure.
4642 29. Contrary to the C standard, it seems that in some environments, the
4643 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4644 program. Exim now does this explicitly; it affects the formatting of
4645 timestamps using strftime().
4647 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4648 complaints. I've now initialized all the variables, to avoid this.
4650 32. Header lines added by a system filter were not being "seen" during
4651 transport-time rewrites.
4653 33. The info_callback() function passed to OpenSSL is set up with type void
4654 (*)(SSL *, int, int), as described somewhere. However, when calling the
4655 function (actually a macro) that sets it up, the type void(*)() is
4656 expected. I've put in a cast to prevent warnings from picky compilers.
4658 34. If a DNS black list lookup found a CNAME record, but there were no A
4659 records associated with the domain it pointed at, Exim crashed.
4661 35. If a DNS black list lookup returned more than one A record, Exim ignored
4662 all but the first. It now scans all returned addresses if a particular IP
4663 value is being sought. In this situation, the contents of the
4664 $dnslist_value variable are a list of all the addresses, separated by a
4667 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4668 to accept a host name and all its aliases if the forward lookup for any of
4669 them yielded the IP address of the incoming connection. Now it accepts only
4670 those names whose forward lookup yields the correct IP address. Any other
4671 names are discarded. This closes a loophole whereby a rogue DNS
4672 administrator could create reverse DNS records to break through a
4673 wildcarded host restriction in an ACL.
4675 37. If a user filter or a system filter that ran in a subprocess used any of
4676 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4677 the wrong values were passed to the pipe command ($thisaddress had the
4678 value of $0, $0 had the value of $1, etc). This bug was introduced by
4679 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4681 38. Improved the line breaking for long SMTP error messages from ACLs.
4682 Previously, if there was no break point between 40 and 75 characters, Exim
4683 left the rest of the message alone. Two changes have been made: (a) I've
4684 reduced the minimum length to 35 characters; (b) if it can't find a break
4685 point between 35 and 75 characters, it looks ahead and uses the first one
4686 that it finds. This may give the occasional overlong line, but at least the
4687 remaining text gets split now.
4689 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4690 file descriptors might be low, and that setting 1000 would always raise it.
4691 It turns out that in some environments, the limit is already over 1000 and
4692 that lowering it causes trouble. So now Exim takes care not to decrease it.
4694 40. When delivering a message, the value of $return_path is set to $sender_
4695 address at the start of routing (routers may change the value). By an
4696 oversight, this default was not being set up when an address was tested by
4697 -bt or -bv, which affected the outcome if any router or filter referred to
4700 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4701 log only when "message" or "log_message" are set. However, if one of the
4702 conditions was an address verification, or a call to a nested ACL, the
4703 messages generated by the underlying test were being passed through. This
4704 no longer happens. The underlying message is available in $acl_verify_
4705 message for both "message" and "log_message" expansions, so it can be
4706 passed through if needed.
4708 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4709 new expansion $bh_ to give the encoded byte string without charset
4710 translation. Translation happens only if iconv() is available; HAVE_ICONV
4711 indicates this at build time. HEADERS_CHARSET gives the charset to
4712 translate to; headers_charset can change it in the configuration, and
4713 "headers charset" can change it in an individual filter file.
4715 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4716 that creates RFC 2047 encoded "words" labels them as that charset instead
4717 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4718 expansion operator; (ii) when Exim creates a From: line for a local
4719 message; (iii) when a header line is rewritten to include a "phrase" part.
4721 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4722 buggy, causing it to skip the first lines of messages whose message ID
4723 ended in 'D'. This would not have bitten before Exim release 4.14, because
4724 message IDs were unlikely to end in 'D' before then. The effect was to have
4725 incorrect size information for certain domains.
4727 45. #include "config.h" was missing at the start of the crypt16.c module. This
4728 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4731 46. If there was a timeout during a "random" callout check, Exim treated it as
4732 a failure of the random address, and carried on sending RSET and the real
4733 address. If the delay was just some slowness somewhere, the response to the
4734 original RCPT would be taken as a response to RSET and so on, causing
4735 mayhem of various kinds.
4737 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4738 when I implemented it. It didn't allow for the fact that some option values
4739 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4740 the right test for positive values.
4742 48. Domain names in DNS records are case-independent. Exim always looks them up
4743 in lower case. Some resolvers return domain names in exactly the case they
4744 appear in the zone file, that is, they may contain uppercase letters. Not
4745 all resolvers do this - some return always lower case. Exim was treating a
4746 change of case by a resolver as a change of domain, similar to a widening
4747 of a domain abbreviation. This triggered its re-routing code and so it was
4748 trying to route what was effectively the same domain again. This normally
4749 caused routing to fail (because the router wouldn't handle the domain
4750 twice). Now Exim checks for this case specially, and just changes the
4751 casing of the domain that it ultimately uses when it transmits the message
4754 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4757 50. If a filter generated a file delivery with a non-absolute name (possible if
4758 no home directory exists for the router), the forbid_file option was not
4761 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4762 the existing equality matching.
4764 52. Exim was using ints instead of ino_t variables in some places where it was
4765 dealing with inode numbers.
4767 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4768 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4769 called TMPDIR, and if it finds it is different, it changes its value.
4771 54. The smtp_printf() function is now made available to local_scan() so
4772 additional output lines can be written before returning. There is also an
4773 smtp_fflush() function to enable the detection of a dropped connection.
4774 The variables smtp_input and smtp_batched_input are exported to
4777 55. Changed the default runtime configuration: the message "Unknown user"
4778 has been removed from the ACL, and instead placed on the localuser router,
4779 using the cannot_route_message feature. This means that any verification
4780 failures that generate their own messages won't get overridden. Similarly,
4781 the "Unrouteable address" message that was in the ACL for unverifiable
4782 relay addresses has also been removed.
4784 56. Added hosts_avoid_esmtp to the smtp transport.
4786 57. The exicyclog script was not checking for the esoteric option
4787 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4788 will work only if exicyclog is run under the appropriate euid.
4790 58. Following a discussion on the list, the rules by which Exim recognises line
4791 endings on incoming messages have been changed. The -dropcr and drop_cr
4792 options are now no-ops, retained only for backwards compatibility. The
4793 following line terminators are recognized: LF CRLF CR. However, special
4794 processing applies to CR:
4796 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4797 nor a local message in the state where . is a terminator.
4799 (ii) If a bare CR is encountered in a header line, an extra space is added
4800 after the line terminator so as not to end the header. The reasoning
4801 behind this is that bare CRs in header lines are most likely either
4802 to be mistakes, or people trying to play silly games.
4804 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4805 was being incorrectly given as 18 bytes larger than it should have been.
4806 This is a VOB (very old bug).
4808 60. This may never have affected anything current, but just in case it has:
4809 When the local host is found other than at the start of a list of hosts,
4810 the local host, those with the same MX, and any that follow, are discarded.
4811 When the list in question was part of a longer list of hosts, the following
4812 hosts (not currently being processed) were also being discarded. This no
4813 longer happens. I'm not sure if this situation could ever has previously
4816 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4819 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4820 header. This is something that is recommended in a new Internet Draft, and
4821 is something that is documented as being done by Sendmail. There are two
4822 possible values. For messages generated by the autoreply transport, Exim
4825 Auto-Submitted: auto-replied
4827 whereas for all other generated messages (e.g. bounces) it adds
4829 Auto-Submitted: auto-generated
4831 63. The "personal" condition in filters now includes a test for the
4832 Auto-Submitted: header. If it contains the string "auto-" the message it
4833 not considered personal.
4835 64. Added rcpt_include_affixes as a generic transport option.
4837 65. Added queue_only_override (default true).
4839 66. Added the syslog_duplication option.
4841 67. If what should have been the first header line of a message consisted of
4842 a space followed by a colon, Exim was mis-interpreting it as a header line.
4843 It isn't of course - it is syntactically invalid and should therefore be
4844 treated as the start of the message body. The misbehaviour could have
4845 caused a number of strange effects, including loss of data in subsequent
4846 header lines, and spool format errors.
4848 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4849 client host had authenticated. This control can now be exercised by an ACL
4850 for more flexibility.
4852 69. By default, callouts do not happen when testing with -bh. There is now a
4853 variant, -bhc, which does actually run the callout code, including
4854 consulting and updating the callout cache.
4856 70. Added support for saslauthd authentication, courtesy of Alexander
4859 71. If statvfs() failed on the spool or log directories while checking their
4860 size for availability, Exim confusingly gave the error "space shortage".
4861 Furthermore, in debugging mode it crashed with a floating point exception.
4862 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4863 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4864 serious problem, Exim now writes to the main and panic logs when this
4865 happens, with details of the failure. It then refuses to accept the
4866 incoming message, giving the message "spool directory problem" or "log
4867 directory problem" with a 421 code for SMTP messages.
4869 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4870 0, 1, and 2 exist, because some OS complain for execs without them (see
4871 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4872 descriptors. However, the code omitted to check that the open succeeded,
4873 causing mysterious errors if for some reason the permissions on /dev/null
4874 got screwed. Now Exim writes a message to the main and panic logs, and
4875 bombs out if it can't open /dev/null.
4877 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4878 interact so that it is all more flexible. It is supposed to remain
4879 backwards compatible. Also added extra_local_interfaces.
4881 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4882 to bomb out with an assertion failure - to the client this appears as a
4883 connection drop. This problem occurs in the part of the code that was taken
4884 from the Samba project. Fortunately, the assertion is in a very simple
4885 function, so I have fixed this by reproducing the function inline in the
4886 one place where it is called, and arranging for authentication to fail
4887 instead of killing the process with assert().
4889 75. The SPA client code was not working when the server requested OEM rather
4890 than Unicode encoding.
4892 76. Added code to make require_files with a specific uid setting more usable in
4893 the case where statting the file as root fails - usually a non-root-mounted
4894 NFS file system. When this happens and the failure is EACCES, Exim now
4895 forks a subprocess and does the per-uid checking as the relevant uid.
4897 77. Added process_log_path.
4899 78. If log_file_path was not explicitly set, a setting of check_log_space or
4900 check_log_inodes was ignored.
4902 79. If a space check for the spool or log partitions fails, the incident is now
4903 logged. Of course, in the latter case the data may get lost...
4905 80. Added the %p formatting code to string_format() so that it can be used to
4906 print addresses in debug_print(). Adjusted all the address printing in the
4907 debugging in store.c to use %p rather than %d.
4909 81. There was a concern that a line of code in smtp_in.c could overflow a
4910 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4911 initially expressed, the concern was not well-founded, because trailing
4912 spaces are removed early. However, if the trailing spaces were followed by
4913 a NULL, they did not get removed, so the overflow was possible. Two fixes
4916 (a) I re-wrote the offending code in a cleaner fashion.
4917 (b) If an incoming SMTP command contains a NULL character, it is rejected
4920 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4921 runs initgroups(), so that if the Exim user is in any additional groups,
4922 they will be used during message reception.
4928 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob