1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.551 2008/08/07 11:05:03 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.
69 TK/01 Add preliminary DKIM support. Currently requires a forked version of
70 ALT-N's libdkim that I have put here:
71 http://duncanthrax.net/exim-experimental/
73 Note to Michael Haardt: I had to rename some vars in sieve.c. They
74 were called 'true' and it seems that C99 defines that as a reserved
75 keyword to be used with 'bool' variable types. That means you could
76 not include C99-style headers which use bools without triggering
77 build errors in sieve.c.
79 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
80 as mailq or other aliases. Changed the --help handling significantly
81 to do whats expected. exim_usage() emits usage/help information.
83 SC/01 Added the -bylocaldomain option to eximstats.
85 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
87 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
89 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
95 PH/01 Another patch from the Sieve maintainer.
97 PH/02 When an IPv6 address is converted to a string for single-key lookup
98 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
99 dots are used instead of colons so that keys in lsearch files need not
100 contain colons. This was done some time before quoting was made available
101 in lsearch files. However, iplsearch files do require colons in IPv6 keys
102 (notated using the quote facility) so as to distinguish them from IPv4
103 keys. This meant that lookups for IP addresses in host lists did not work
104 for iplsearch lookups.
106 This has been fixed by arranging for IPv6 addresses to be expressed with
107 colons if the lookup type is iplsearch. This is not incompatible, because
108 previously such lookups could never work.
110 The situation is now rather anomolous, since one *can* have colons in
111 ordinary lsearch keys. However, making the change in all cases is
112 incompatible and would probably break a number of configurations.
114 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
117 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
118 conversion specification without a maximum field width, thereby enabling
119 a rogue spamd server to cause a buffer overflow. While nobody in their
120 right mind would setup Exim to query an untrusted spamd server, an
121 attacker that gains access to a server running spamd could potentially
122 exploit this vulnerability to run arbitrary code as the Exim user.
124 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
125 $primary_hostname instead of what libspf2 thinks the hosts name is.
127 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
128 a directory entry by the name of the lookup key. Previously, if a
129 symlink pointed to a non-existing file or a file in a directory that
130 Exim lacked permissions to read, a lookup for a key matching that
131 symlink would fail. Now it is enough that a matching directory entry
132 exists, symlink or not. (Bugzilla 503.)
134 PH/03 The body_linecount and body_zerocount variables are now exported in the
137 PH/04 Added the $dnslist_matched variable.
139 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
140 This means they are set thereafter only if the connection becomes
143 PH/06 Added the client_condition to authenticators so that some can be skipped
144 by clients under certain conditions.
146 PH/07 The error message for a badly-placed control=no_multiline_responses left
147 "_responses" off the end of the name.
149 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
151 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
152 (without spaces) instead of just copying the configuration text.
154 PH/10 Added the /noupdate option to the ratelimit ACL condition.
156 PH/11 Added $max_received_linelength.
158 PH/12 Added +ignore_defer and +include_defer to host lists.
160 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
161 way in which PCRE > 7.0 is built.
163 PH/14 Implemented queue_only_load_latch.
165 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
166 MAIL command. The effect was to mangle the value on 64-bit systems.
168 PH/16 Another patch from the Sieve maintainer.
170 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
172 PH/18 If a system quota error occurred while trying to create the file for
173 a maildir delivery, the message "Mailbox is full" was not appended to the
174 bounce if the delivery eventually timed out. Change 4.67/27 below applied
175 only to a quota excession during the actual writing of the file.
177 PH/19 It seems that peer DN values may contain newlines (and other non-printing
178 characters?) which causes problems in log lines. The DN values are now
179 passed through string_printing() before being added to log lines.
181 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
182 and InterBase are left for another time.)
184 PH/21 Added message_body_newlines option.
186 PH/22 Guard against possible overflow in moan_check_errorcopy().
188 PH/23 POSIX allows open() to be a macro; guard against that.
190 PH/24 If the recipient of an error message contained an @ in the local part
191 (suitably quoted, of course), incorrect values were put in $domain and
192 $local_part during the evaluation of errors_copy.
198 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
199 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
202 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
203 issue a MAIL command.
205 PH/02 In an ACL statement such as
207 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
209 if a client was not listed at all, or was listed with a value other than
210 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
211 the condition was not true (as it should be), so access was not denied.
212 The bug was that the ! inversion was incorrectly passed on to the second
213 item. This has been fixed.
215 PH/03 Added additional dnslists conditions == and =& which are different from
216 = and & when the dns lookup returns more than one IP address.
218 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
219 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
221 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
222 FSYNC, which compiles an option called disable_fsync that allows for
223 bypassing fsync(). The documentation is heavily laced with warnings.
225 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
227 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
228 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
229 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
230 including adding "make clean"; (3) Added -fPIC when compiling the test
231 dynamically loaded module, to get rid of a warning.
233 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
234 message fails, move_frozen_messages = true and ignore_bounce_errors_after
235 = 0s. The bug is otherwise harmless.
237 PH/07 There was a bug in the dovecot authenticator such that the value of
238 $auth1 could be overwritten, and so not correctly preserved, after a
239 successful authentication. This usually meant that the value preserved by
240 the server_setid option was incorrect.
242 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
244 PH/09 Installed PCRE release 7.0.
246 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
247 run for batched SMTP input. It is now run at the start of every message
248 in the batch. While fixing this I discovered that the process information
249 (output by running exiwhat) was not always getting set for -bs and -bS
250 input. This is fixed, and it now also says "batched" for BSMTP.
252 PH/11 Added control=no_pipelining.
254 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
255 patch, slightly modified), and move the expansion of helo_data till after
256 the connection is made in the smtp transport (so it can use these
259 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
261 PH/14 Added log_selector = +pid.
263 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
265 PH/16 Add ${if forany and ${if forall.
267 PH/17 Added dsn_from option to vary the From: line in DSNs.
269 PH/18 Flush SMTP output before performing a callout, unless control =
270 no_callout_flush is set.
272 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
273 was true (the default) a successful delivery failed to delete the retry
274 item, thus causing premature timeout of the address. The bug is now
277 PH/20 Added hosts_avoid_pipelining to the smtp transport.
279 PH/21 Long custom messages for fakedefer and fakereject are now split up
280 into multiline reponses in the same way that messages for "deny" and
281 other ACL rejections are.
283 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
284 with slight modification.
286 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
287 draft, changing the syntax and factoring some duplicate code".
289 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
290 for deliveries of the second and subsequent messages over the same SMTP
293 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
294 ${reduce, with only minor "tidies".
296 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
298 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
299 expansion side effects.
301 PH/27 When a message times out after an over-quota error from an Exim-imposed
302 quota, the bounce message says "mailbox is full". This message was not
303 being given when it was a system quota that was exceeded. It now should
306 MH/03 Made $recipients available in local_scan(). local_scan() already has
307 better access to the recipient list through recipients_list[], but
308 $recipients can be useful in postmaster-provided expansion strings.
310 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
311 in the case of a MAIL command with additional options following the
312 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
313 were accidentally chopped off.
315 PH/29 SMTP synchronization checks are implemented when a command is read -
316 there is a check that no more input is waiting when there shouldn't be
317 any. However, for some commands, a delay in an ACL can mean that it is
318 some time before the response is written. In this time, more input might
319 arrive, invalidly. So now there are extra checks after an ACL has run for
320 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
321 pipelining has not been advertised.
323 PH/30 MH's patch to allow iscntrl() characters to be list separators.
325 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
326 returned in the SMTP response when smtp_return_error_details was false.
329 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
330 instead of the C I/O that was originally supplied, because problems were
333 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
334 Exim which did not show up earlier: it was assuming that a call to
335 SSL_CTX_set_info_callback() might give an error value. In fact, there is
336 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
337 was a macro that became an assignment, so it seemed to work. This has
338 changed to a proper function call with a void return, hence the compile
339 error. Exim's code has been fixed.
341 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
344 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
346 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
348 PH/37 If a message is not accepted after it has had an id assigned (e.g.
349 because it turns out to be too big or there is a timeout) there is no
350 "Completed" line in the log. When some messages of this type were
351 selected by exigrep, they were listed as "not completed". Others were
352 picked up by some special patterns. I have improved the selection
353 criteria to be more general.
355 PH/38 The host_find_failed option in the manualroute router can now be set
356 to "ignore", to completely ignore a host whose IP address cannot be
357 found. If all hosts are ignored, the behaviour is controlled by the new
358 host_all_ignored option.
360 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
361 homing or because of multiple MX records with /mx) generated more than
362 one IP address, and the following item turned out to be the local host,
363 all the secondary addresses of the first item were incorrectly removed
364 from the list, along with the local host and any following hosts (which
365 is what is supposed to happen).
367 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
368 whoever called Exim into the -H file. In the case of the daemon it was
369 behaving confusingly. When first started, it used values for whoever
370 started the daemon, but after a SIGHUP it used the Exim user (because it
371 calls itself on a restart). I have changed the code so that it now always
374 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
375 message are rejected with the same error (e.g. no authentication or bad
376 sender address), and a DATA command is nevertheless sent (as can happen
377 with PIPELINING or a stupid MUA), the error message that was given to the
378 RCPT commands is included in the rejection of the DATA command. This is
379 intended to be helpful for MUAs that show only the final error to their
382 PH/42 Another patch from the Sieve maintainer.
384 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
385 Eximstats - Fixed some broken HTML links and added missing column headers
387 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
388 columns for Rejects, Temp Rejects, Ham, and Spam rows.
390 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
392 PH/43 Yet another patch from the Sieve maintainer.
394 PH/44 I found a way to check for a TCP/IP connection going away before sending
395 the response to the final '.' that terminates a message, but only in the
396 case where the client has not sent further data following the '.'
397 (unfortunately, this is allowed). However, in many cases there won't be
398 any further data because there won't be any more messages to send. A call
399 to select() can be used: if it shows that the input is "ready", there is
400 either input waiting, or the socket has been closed. An attempt to read
401 the next input character can distinguish the two cases. Previously, Exim
402 would have sent an OK response which the client would never have see.
403 This could lead to message repetition. This fix should cure that, at
404 least in a lot of common cases.
406 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
407 advertised in response to EHLO.
413 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
414 fixed by 4.65/MH/01 (is this a record?) are fixed:
416 (i) An empty string was always treated as zero by the numeric comparison
417 operators. This behaviour has been restored.
419 (ii) It is documented that the numeric comparison operators always treat
420 their arguments as decimal numbers. This was broken in that numbers
421 starting with 0 were being interpreted as octal.
423 While fixing these problems I realized that there was another issue that
424 hadn't been noticed. Values of message_size_limit (both the global option
425 and the transport option) were treated as octal if they started with 0.
426 The documentation was vague. These values are now always treated as
427 decimal, and I will make that clear in the documentation.
433 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
434 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
437 MH/01 Don't check that the operands of numeric comparison operators are
438 integers when their expansion is in "skipping" mode (fixes bug
439 introduced by 4.64-PH/07).
441 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
442 child addresses, Exim now panics and dies. Previously, because the count
443 is held in a short int, deliveries were likely to be lost. As such a
444 large number of recipients for a single message is ridiculous
445 (performance will be very, very poor), I have chosen to impose a limit
446 rather than extend the field.
452 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
453 leftover -K file (the existence of which was triggered by #402).
454 While we were at it, introduced process PID as part of the -K
455 filename. This should rule out race conditions when creating
458 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
459 processing considerably. Previous code took too long for large mails,
460 triggering a timeout which in turn triggers #401.
462 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
463 in the DK code in transports.c. sendfile() is not really portable,
464 hence the _LINUX specificness.
466 TF/01 In the add_headers option to the mail command in an Exim filter,
467 there was a bug that Exim would claim a syntax error in any
468 header after the first one which had an odd number of characters
471 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
472 callout verification, Exim cached a "reject" for the entire domain. This
473 is correct for most verifications, but it is not correct for a recipient
474 verification with use_sender or use_postmaster set, because in that case
475 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
476 case of MAIL FROM:<> rejection from other early rejections (e.g.
477 rejection of HELO). When verifying a recipient using a non-null MAIL
478 address, the cache is ignored if it shows MAIL FROM:<> rejection.
479 Whatever the result of the callout, the value of the domain cache is
480 left unchanged (for any other kind of callout, getting as far as trying
481 RCPT means that the domain itself is ok).
483 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
486 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
487 manifest itself as EPIPE rather than ECONNECT. When tidying away a
488 session, the daemon ignores ECONNECT errors and logs others; it now
489 ignores EPIPE as well.
491 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
492 (quoted-printable decoding).
494 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
495 later the small subsequent patch to fix an introduced bug.
497 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
499 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
501 PH/08 An error is now given if message_size_limit is specified negative.
503 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
504 to be given (somewhat) arbitrary names.
506 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
509 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
510 miscellaneous code fixes
512 PH/10 Added the log_reject_target ACL modifier to specify where to log
515 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
516 hostname. This is wrong, because it relates to the incoming message (and
517 probably the interface on which it is arriving) and not to the outgoing
518 callout (which could be using a different interface). This has been
519 changed to use the value of the helo_data option from the smtp transport
520 instead - this is what is used when a message is actually being sent. If
521 there is no remote transport (possible with a router that sets up host
522 addresses), $smtp_active_hostname is used.
524 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
525 tweaks were necessary in order to get it to work (see also 21 below):
526 (a) The code assumed that strncpy() returns a negative number on buffer
527 overflow, which isn't the case. Replaced with Exim's string_format()
529 (b) There were several signed/unsigned issues. I just did the minimum
530 hacking in of casts. There is scope for a larger refactoring.
531 (c) The code used strcasecmp() which is not a standard C function.
532 Replaced with Exim's strcmpic() function.
533 (d) The code set only $1; it now sets $auth1 as well.
534 (e) A simple test gave the error "authentication client didn't specify
535 service in request". It would seem that Dovecot has changed its
536 interface. Fortunately there's a specification; I followed it and
537 changed what the client sends and it appears to be working now.
539 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
542 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
543 address A is aliased to B and C, where B exists and C does not. Without
544 -v the output is "A verified" because verification stops after a
545 successful redirection if more than one address is generated. However,
546 with -v the child addresses are also verified. Exim was outputting "A
547 failed to verify" and then showing the successful verification for C,
548 with its parentage. It now outputs "B failed to verify", showing B's
549 parentage before showing the successful verification of C.
551 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
552 look up a TXT record in a specific list after matching in a combined
555 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
556 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
557 they consult the DNS. I had assumed they would set it the way they
558 wanted; and indeed my experiments on Linux seem to show that in some
559 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
560 To be on the safe side, however, I have now made the interface to
561 host_find_byname() similar to host_find_bydns(), with an argument
562 containing the DNS resolver options. The host_find_byname() function now
563 sets these options at its start, just as host_find_bydns() does. The smtp
564 transport options dns_qualify_single and dns_search_parents are passed to
565 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
566 of host_find_byname() use the default settings of RES_DEFNAMES
567 (qualify_single) but not RES_DNSRCH (search_parents).
569 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
570 spool_read_header() do less string testing, by means of a preliminary
571 switch on the second character of optional "-foo" lines. (This is
572 overdue, caused by the large number of possibilities that now exist.
573 Originally there were few.) While I was there, I also converted the
574 str(n)cmp tests so they don't re-test the leading "-" and the first
575 character, in the hope this might squeeze out yet more improvement.
577 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
578 flag allowing group syntax was set by the header_syntax check but not
579 turned off, possible causing trouble later; (2) The flag was not being
580 set at all for the header_verify test, causing "group"-style headers to
581 be rejected. I have now set it in this case, and also caused header_
582 verify to ignore an empty address taken from a group. While doing this, I
583 came across some other cases where the code for allowing group syntax
584 while scanning a header line wasn't quite right (mostly, not resetting
585 the flag correctly in the right place). These bugs could have caused
586 trouble for malformed header lines. I hope it is now all correct.
588 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
589 with the "reply" argument non-NULL. The code, however (which originally
590 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
591 but it didn't always do it. This confused somebody who was copying the
592 code for some other use. I have removed all the tests.
594 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
595 feature that was used to support insecure browsers during the U.S. crypto
596 embargo. It requires special client support, and Exim is probably the
597 only MTA that supported it -- and would never use it because real RSA is
598 always available. This code has been removed, because it had the bad
599 effect of slowing Exim down by computing (never used) parameters for the
600 RSA_EXPORT functionality.
602 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
603 authenticator to fail if there's a tab character in the incoming data
604 (there should never be unless someone is messing about, as it's supposed
605 to be base64-encoded). Also added, on Timo's advice, the "secured" option
606 if the connection is using TLS or if the remote IP is the same as the
607 local IP, and the "valid-client-cert option" if a client certificate has
610 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
611 authenticators. This can be used for authorization after authentication
612 succeeds. (In the case of plaintext, it servers for both authentication
615 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
616 if any retry times were supplied.
618 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
619 connection, where there is no client IP address to check. In this
620 situation, the verify now always succeeds.
622 PH/25 Applied John Jetmore's -Mset patch.
624 PH/26 Added -bem to be like -Mset, but loading a message from a file.
626 PH/27 In a string expansion for a processed (not raw) header when multiple
627 headers of the same name were present, leading whitespace was being
628 removed from all of them, but trailing whitespace was being removed only
629 from the last one. Now trailing whitespace is removed from each header
630 before concatenation. Completely empty headers in a concatenation (as
633 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
634 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
636 PH/29 [Removed. This was a change that I later backed out, and forgot to
637 correct the ChangeLog entry (that I had efficiently created) before
638 committing the later change.]
640 PH/30 Exim was sometimes attempting to deliver messages that had suffered
641 address errors (4xx response to RCPT) over the same connection as other
642 messages routed to the same hosts. Such deliveries are always "forced",
643 so retry times are not inspected. This resulted in far too many retries
644 for the affected addresses. The effect occurred only when there were more
645 hosts than the hosts_max_try setting in the smtp transport when it had
646 the 4xx errors. Those hosts that it had tried were not added to the list
647 of hosts for which the message was waiting, so if all were tried, there
648 was no problem. Two fixes have been applied:
650 (i) If there are any address or message errors in an SMTP delivery, none
651 of the hosts (tried or untried) are now added to the list of hosts
652 for which the message is waiting, so the message should not be a
653 candidate for sending over the same connection that was used for a
654 successful delivery of some other message. This seems entirely
655 reasonable: after all the message is NOT "waiting for some host".
656 This is so "obvious" that I'm not sure why it wasn't done
657 previously. Hope I haven't missed anything, but it can't do any
658 harm, as the worst effect is to miss an optimization.
660 (ii) If, despite (i), such a delivery is accidentally attempted, the
661 routing retry time is respected, so at least it doesn't keep
662 hammering the server.
664 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
665 in ${readsocket because some servers need this prod.
667 PH/32 Added some extra debug output when updating a wait-xxx database.
669 PH/33 The hint "could be header name not terminated by colon", which has been
670 given for certain expansion errors for a long time, was not being given
671 for the ${if def:h_colon_omitted{... case.
673 PH/34 The spec says: "With one important exception, whenever a domain list is
674 being scanned, $domain contains the subject domain." There was at least
675 one case where this was not true.
677 PH/35 The error "getsockname() failed: connection reset by peer" was being
678 written to the panic log as well as the main log, but it isn't really
679 panic-worthy as it just means the connection died rather early on. I have
680 removed the panic log writing for the ECONNRESET error when getsockname()
683 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
684 runs only) independently of the message's sender address. This meant
685 that, if the 4xx error was in fact related to the sender, a different
686 message to the same recipient with a different sender could confuse
687 things. In particualar, this can happen when sending to a greylisting
688 server, but other circumstances could also provoke similar problems.
689 I have changed the default so that the retry time for these errors is now
690 based a combination of the sender and recipient addresses. This change
691 can be overridden by setting address_retry_include_sender=false in the
694 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
695 remote server are returned as part of bounce messages. This was not
696 happening for LMTP over a pipe (the lmtp transport), but now it is the
697 same for both kinds of LMTP.
699 PH/38 Despite being documented as not happening, Exim was rewriting addresses
700 in header lines that were in fact CNAMEs. This is no longer the case.
702 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
703 and queue runs started by the daemon processed all messages. This has
704 been fixed so that -R and -S can now usefully be given with -q<time>.
706 PH/40 Import PCRE release 6.7 (fixes some bugs).
708 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
710 PH/42 Give an error if -q is specified more than once.
712 PH/43 Renamed the variables $interface_address and $interface_port as
713 $received_ip_address and $received_port, to make it clear that these
714 values apply to message reception, and not to the outgoing interface when
715 a message is delivered. (The old names remain recognized, of course.)
717 PH/44 There was no timeout on the connect() call when using a Unix domain
718 socket in the ${readsocket expansion. There now is.
720 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
721 be meaningful with "accept".
723 SC/01 Eximstats V1.43
724 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
726 SC/02 Eximstats V1.44
727 Use a glob alias rather than an array ref in the generated
728 parser. This improves both readability and performance.
730 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
731 Collect SpamAssassin and rejection statistics.
732 Don't display local sender or destination tables unless
733 there is data to show.
734 Added average volumes into the top table text output.
736 SC/04 Eximstats V1.46
737 Collect data on the number of addresses (recipients)
738 as well as the number of messages.
740 SC/05 Eximstats V1.47
741 Added 'Message too big' to the list of mail rejection
742 reasons (thanks to Marco Gaiarin).
744 SC/06 Eximstats V1.48
745 Mainlog lines which have GMT offsets and are too short to
746 have a flag are now skipped.
748 SC/07 Eximstats V1.49 (Alain Williams)
749 Added the -emptyok flag.
751 SC/08 Eximstats V1.50
752 Fixes for obtaining the IP address from reject messages.
754 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
755 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
756 whitesspace changes from 4.64-PH/27
758 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
761 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
762 are found, allow negative numbers in numeric criteria)
764 JJ/06 exipick.20061117.2, added new $message_body_missing variable
766 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
767 to match changes made in 4.64-PH/43
769 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
771 PH/47 Put in an explicit test for a DNS lookup of an address record where the
772 "domain" is actually an IP address, and force a failure. This locks out
773 those revolvers/nameservers that support "A-for-A" lookups, in
774 contravention of the specifications.
776 PH/48 When a host name was looked up from an IP address, and the subsequent
777 forward lookup of the name timed out, the host name was left in
778 $sender_host_name, contrary to the specification.
780 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
781 restricted to single-key lookups, Exim was not diagnosing an error if
782 * or *@ was used with a query-style lookup.
784 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
786 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
787 long ago, but noone interested enough thought of it. Let's just say that
788 the "1.1" means that there are some new functions that weren't there at
789 some point in the past.
791 PH/51 Error processing for expansion failure of helo_data from an smtp
792 transport during callout processing was broken.
794 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
795 tested/used via the -bh/-bhc/-bs options.
797 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
798 bug, fixed in subsequent PCRE releases).
800 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
801 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
803 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
809 SC/01 Use a glob alias rather than an array ref in eximstats generated
810 parser. This improves both readability and performance.
812 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
813 Don't display local sender or destination tables in eximstats unless
814 there is data to show.
815 Added average volumes into the eximstats top table text output.
817 SC/03 Collect data on the number of addresses (recipients) as well
818 as the number of messages in eximstats.
820 TF/01 Correct an error in the documentation for the redirect router. Exim
821 does (usually) call initgroups() when daemonizing.
823 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
824 with consistent privilege compared to when running as a daemon.
826 TF/03 Note in the spec that $authenticated_id is not set for local
827 submissions from trusted users.
829 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
830 Thanks to Dean Brooks <dean@iglou.com> for the patch.
832 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
833 by adding some example configuration directives to the default
834 configuration file. A little bit of work is required to uncomment the
835 directives and define how usernames and passwords are checked, but
836 there is now a framework to start from.
838 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
839 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
840 without this. I don't know how relevant this is to other LDAP libraries.
842 PH/02 Add the verb name to the "unknown ACL verb" error.
844 PH/03 Magnus Holmgren's patch for filter_prepend_home.
846 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
848 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
849 directory not expanded when it should be if an expanded home directory
850 was set for the address (which is overridden by the transport).
852 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
855 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
856 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
857 because it is too late at that time, and has no effect.
859 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
860 security issue with \' (bugzilla #107). I could not use the
861 PQescapeStringConn() function, because it needs a PGconn value as one of
864 PH/08 When testing addresses using -bt, indicate those final addresses that
865 are duplicates that would not cause an additional delivery. At least one
866 person was confused, thinking that -bt output corresponded to deliveries.
867 (Suppressing duplicates isn't a good idea as you lose the information
868 about possibly different redirections that led to the duplicates.)
870 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
871 systems where poll() doesn't work, in particular OS X.
873 PH/10 Added more information to debugging output for retry time not reached.
875 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
876 operations in malware.c.
878 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
881 PH/13 If write_rejectlog was set false when logging was sent to syslog with
882 syslog_duplication set false, log lines that would normally be written
883 both the the main log and to the reject log were not written to syslog at
886 PH/14 In the default configuration, change the use of "message" in ACL warn
887 statements to "add_header".
889 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
890 not followed by a command (e.g. "seen endif").
892 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
893 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
896 PH/17 Added extra conditions to the default value of delay_warning_condition
900 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
901 { match{$h_precedence:}{(?i)bulk|list|junk} } \
902 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
905 The Auto-Submitted: and various List- headers are standardised, whereas I
906 don't think Precedence: ever was.
908 PH/18 Refactored debugging code in route_finduser() to show more information,
909 in particular, the error code if getpwnam() issues one.
911 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
912 This is apparently needed in addition to the PH/07 change above to avoid
913 any possible encoding problems.
915 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
916 but not after initializing Perl.
918 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
919 output them only if debugging. By default they are written stderr,
920 apparently, which is not desirable.
922 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
925 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
928 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
930 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
931 authenticated or an ident call has been made. Suppress the default
932 values for $authenticated_id and $authenticated_sender (but permit -oMai
933 and -oMas) when testing with -bh.
935 PH/24 Re-jigged the order of the tests in the default configuration so that the
936 tests for valid domains and recipients precede the DNS black list and CSA
937 tests, on the grounds that those ones are more expensive.
939 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
940 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
941 This bug exists in every version of Exim that I still have, right back to
944 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
945 However, an attempt to turn on case-sensitivity in a regex key by
946 including (?-i) didn't work because the subject string was already
947 lowercased, and the effects were non-intuitive. It turns out that a
948 one-line patch can be used to allow (?-i) to work as expected.
954 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
955 other effects) broke the use of negated acl sub-conditions.
957 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
960 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
961 "Deny" causes Exim to reject the incoming connection with a 554 error.
962 Unfortunately, if there is a major crisis, such as a disk failure,
963 tcp-wrappers gives "deny", whereas what one would like would be some
964 kind of temporary error. A kludge has been added to help with this.
965 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
966 554 error is used if errno is still zero or contains ENOENT (which occurs
967 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
970 PH/03 Add -lutil to the default FreeBSD LIBS setting.
972 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
973 errors. Otherwise a message that provokes a temporary error (when other
974 messages do not) can cause a whole host to time out.
976 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
977 addresses were routed directly to files or pipes from a redirect router.
978 File deliveries just didn't batch; pipe deliveries might have suffered
981 PH/06 A failure to get a lock for a hints database would erroneously always say
982 "Failed to get write lock", even when it was really a read lock.
984 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
985 of 0600. This has been changed to use the value of the lockfile_mode
986 option (which defaults to 0600).
988 PH/08 Applied small patch from the Sieve maintainer.
990 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
991 folder from quota calculations, a direct delivery into this folder messed
992 up the contents of the maildirsize file. This was because the regex was
993 used only to exclude .Trash (or whatever) when the size of the mailbox
994 was calculated. There was no check that a delivery was happening into an
995 excluded directory. This bug has been fixed by ignoring all quota
996 processing for deliveries into excluded directories.
998 PH/10 Added the maildirfolder_create_regex option to appendfile.
1004 PH/01 The code for finding all the local interface addresses on a FreeBSD
1005 system running IPv6 was broken. This may well have applied to all BSD
1006 systems, as well as to others that have similar system calls. The broken
1007 code found IPv4 interfaces correctly, but gave incorrect values for the
1008 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1009 that it would not match correctly against @[] and not recognize the IPv6
1012 PH/02 The ipliteral router was not recognizing addresses of the form user@
1013 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1015 PH/03 Added disable_ipv6.
1017 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1018 decoded form, because it is most often used to construct To: headers
1019 lines in autoreplies, and the decoded form may well be syntactically
1020 invalid. However, $reply_address has leading white space removed, and all
1021 newlines turned into spaces so that the autoreply transport does not
1024 PH/05 If group was specified without a user on a router, and no group or user
1025 was specified on a transport, the group from the router was ignored.
1027 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1028 for visible compile-time settings that can be used to change these
1029 numbers, for those that want even more. Backwards compatibility with old
1030 spool files has been maintained. However, going back to a previous Exim
1031 release will lost any variables that are in spool files.
1033 PH/07 Two small changes when running in the test harness: increase delay when
1034 passing a TCP/IP connection to a new process, in case the original
1035 process has to generate a bounce, and remove special handling of
1036 127.0.0.2 (sic), which is no longer necessary.
1038 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1039 be the same on different OS.
1041 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1044 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1045 whether --show-vars was specified or not
1047 JJ/02 exipick: Added support for new ACL variable spool format introduced
1050 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1051 syntactically invalid From: or Reply-to: line, and a filter used this to
1052 generate an autoreply, and therefore failed to obtain an address for the
1053 autoreply, Exim could try to deliver to a non-existent relative file
1054 name, causing unrelated and misleading errors. What now happens is that
1055 it logs this as a hard delivery error, but does not attempt to create a
1058 PH/11 The exinext utility has a -C option for testing purposes, but although
1059 the given file was scanned by exinext itself; it wasn't being passed on
1060 when Exim was called.
1062 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1063 an end-of-file indication when reading a command response.
1065 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1066 compiled. In many other places in Exim, IPv6 addresses are always
1067 recognized, so I have changed this. It also means that IPv4 domain
1068 literals of the form [IPV4:n.n.n.n] are now always recognized.
1070 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1071 used if the router is not running as root, for example, when verifying at
1072 ACL time, or when using -bh. The debugging output from this situation was
1073 non-existent - all you got was a failure to exec. I have made two
1076 (a) Failures to set uid/gid, the current directory, or a process leader
1077 in a subprocess such as that created by queryprogram now generate
1078 suitable debugging ouput when -d is set.
1080 (b) The queryprogram router detects when it is not running as root,
1081 outputs suitable debugging information if -d is set, and then runs
1082 the subprocess without attempting to change uid/gid.
1084 PH/15 Minor change to Makefile for building test_host (undocumented testing
1087 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1088 additional section of a DNS packet that returns MX or SRV records.
1089 Instead, it always explicitly searches for A/AAAA records. This avoids
1090 major problems that occur when a DNS server includes only records of one
1091 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1092 fixed another bug: if SRV records were looked up and the corresponding
1093 address records were *not* found in the additional section, the port
1094 values from the SRV records were lost.
1096 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1097 using the correct key (the original address) when searching the retry
1098 rules in order to find which one to use for generating the retry hint.
1100 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1101 adding the default one. Similarly, if it contains a Reply-To: header, the
1102 errors_reply_to option, if set, is not used.
1104 PH/19 When calculating a retry time, Exim used to measure the "time since
1105 failure" by looking at the "first failed" field in the retry record. Now
1106 it does not use this if it is later than than the arrival time of the
1107 message. Instead it uses the arrival time. This makes for better
1108 behaviour in cases where some deliveries succeed, thus re-setting the
1109 "first failed" field. An example is a quota failure for a huge message
1110 when small messages continue to be delivered. Without this change, the
1111 "time since failure" will always be short, possible causing more frequent
1112 delivery attempts for the huge message than are intended.
1113 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1115 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1116 $1, $2, $3) because the numerical variables can be reset during some
1117 expansion items (e.g. "match"), thereby losing the authentication data.
1119 PH/21 Make -bV show the size of off_t variables so that the test suite can
1120 decide whether to run tests for quotas > 2G.
1122 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1123 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1124 filecount value is greater than 2G or if a quota value is greater than 2G
1125 on a system where the size of off_t is not greater than 4, a panic error
1128 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1129 never match. The debug and -bh output now contains an explicit error
1130 message indicating a malformed IPv4 address or mask.
1132 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1133 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1134 PH/23 above applies.
1136 PH/25 Do not write to syslog when running in the test harness. The only
1137 occasion when this arises is a failure to open the main or panic logs
1138 (for which there is an explicit test).
1140 PH/26 Added the /no_tell option to "control=freeze".
1142 PH/27 If a host name lookup failed very early in a connection, for example, if
1143 the IP address matched host_lookup and the reverse lookup yielded a name
1144 that did not have a forward lookup, an error message of the form "no IP
1145 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1146 could be logged. Now it outputs the IP address instead of "NULL".
1148 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1149 allows the sender and the authenticated sender to be set when
1150 submitting a message from within Exim. Since child_open_exim() is
1151 documented for local_scan(), the new function should be too.
1153 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1154 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1155 results in an empty string is now treated as unset.
1157 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1159 PH/31 Added sender_verify_fail logging option.
1161 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1162 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1163 tidied the source and removed it altogether.
1165 PH/33 When a queue run was abandoned because the load average was too high, a
1166 log line was always written; now it is written only if the queue_run log
1167 selector is set. In addition, the log line for abandonment now contains
1168 information about the queue run such as the pid. This is always present
1169 in "start" and "stop" lines but was omitted from the "abandon" line.
1171 PH/34 Omit spaces between a header name and the colon in the error message that
1172 is given when verify = headers_syntax fails (if there are lots of them,
1173 the message gets confusing).
1175 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1176 names, as there are now some PTR records that contain slashes. This check
1177 is only to protect against broken name servers that fall over on strange
1178 characters, so the fact that it applies to all lookups doesn't matter.
1180 PH/36 Now that the new test suite is complete, we can remove some of the
1181 special code in Exim that was needed for the old test suite. For example,
1182 sorting DNS records because real resolvers return them in an arbitrary
1183 order. The new test suite's fake resolver always returns records in the
1186 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1187 bounces) except when queue_only is set, to avoid logging races between
1188 the different processes.
1190 PH/38 Panic-die if .include specifies a non-absolute path.
1192 PH/39 A tweak to the "H" retry rule from its user.
1194 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1195 a label. They prevented compilation on older perls.
1197 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1198 a warning to be raised on newish perls.
1200 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1201 on queue. Changes to match documented behaviour of showing count of
1202 messages matching specified criteria.
1204 PH/40 Changed the default ident timeout from 30s to 5s.
1206 PH/41 Added support for the use of login_cap features, on those BSD systems
1207 that have them, for controlling the resources used by pipe deliveries.
1209 PH/42 The content-scanning code uses fopen() to create files in which to put
1210 message data. Previously it was not paying any attention to the mode of
1211 the files. Exim runs with umask(0) because the rest of the code creates
1212 files with open(), and sets the required mode explicitly. Thus, these
1213 files were ending up world-writeable. This was not a big issue, because,
1214 being within the spool directory, they were not world-accessible. I have
1215 created a function called modefopen, which takes an additional mode
1216 argument. It sets umask(777), creates the file, chmods it to the required
1217 mode, then resets the umask. All the relevant calls to fopen() in the
1218 content scanning code have been changed to use this function.
1220 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1221 to 24 hours. This avoids potential overflow problems when processing G
1222 and H retry rules. I suspect nobody ever tinkers with this value.
1224 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1226 PH/45 When the plaintext authenticator is running as a client, the server's
1227 challenges are checked to ensure they are valid base64 strings. By
1228 default, the authentication attempt is cancelled if an invalid string is
1229 received. Setting client_ignore_invalid_base64 true ignores these errors.
1230 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1231 they are received. Thus, the responses can be made to depend on the
1232 challenges. If an invalid string is ignored, an empty string is placed in
1235 PH/46 Messages that are created by the autoreply transport now contains a
1236 References: header, in accordance with RFCs 2822 and 3834.
1238 PH/47 Added authenticated_sender_force to the smtp transport.
1240 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1242 PH/49 Installed latest patch from the Sieve maintainer.
1244 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1245 was also set, the appendfile transport was unnecessarily scanning a
1246 directory of message files (e.g. for maildir delivery) to find the count
1247 of files (along with the size), even though it did not need this
1248 information. It now does the scan only if it needs to find either the
1249 size of the count of files.
1251 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1253 PH/52 Two bugs concerned with error handling when the smtp transport is
1256 (i) Exim was not creating retry information for temporary errors given
1257 for individual recipients after the DATA command when the smtp transport
1258 was used in LMTP mode. This meant that they could be retried too
1259 frequently, and not timed out correctly.
1261 (ii) Exim was setting the flag that allows error details to be returned
1262 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1263 recipients that were returned after the DATA command.
1265 PH/53 This is related to PH/52, but is more general: for any failing address,
1266 when detailed error information was permitted to be returned to the
1267 sender, but the error was temporary, then after the final timeout, only
1268 "retry timeout exceeded" was returned. Now it returns the full error as
1269 well as "retry timeout exceeded".
1271 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1272 do this, and (what is worse) MTAs that accept it.
1274 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1275 will now be deprecated.
1277 PH/56 New os.c-cygwin from the Cygwin maintainer.
1279 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1280 formats (previously only available in exim formats via -bpr, -bpru,
1281 and -bpra. Now also available in native and exiqgrep formats)
1283 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1284 with very large, slow to parse queues
1286 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1288 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1290 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1291 responses to authentication challenges, though it was showing the
1292 challenges; (ii) I've removed the CR characters from the debug output for
1295 PH/58 Allow for the insertion of a newline as well as a space when a string
1296 is turned into more than one encoded-word during RFC 2047 encoding. The
1297 Sieve code now uses this.
1299 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1300 data_4xx, lost_connection, tls_required.
1302 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1303 message was being sent as an SMTP response.
1305 PH/61 Add -l and -k options to exicyclog.
1307 PH/62 When verifying, if an address was redirected to one new address, so that
1308 verification continued, and the new address failed or deferred after
1309 having set something in $address_data, the value of $address_data was not
1310 passed back to the ACL. This was different to the case when no
1311 redirection occurred. The value is now passed back in both cases.
1313 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1314 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1315 use login_cap.h, so on its own it isn't the distinguishing feature. The
1316 new name refers directly to the setclassresources() function.
1318 PH/65 Added configuration files for NetBSD3.
1320 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1322 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1323 is preferred over IPv4.
1325 PH/68 The bounce_return_message and bounce_return_body options were not being
1326 honoured for bounces generated during the reception of non-SMTP messages.
1327 In particular, this applied to messages rejected by the ACL. This bug has
1328 been fixed. However, if bounce_return_message is true and bounce_return_
1329 body is false, the headers that are returned for a non-SMTP message
1330 include only those that have been read before the error was detected.
1331 (In the case of an ACL rejection, they have all been read.)
1333 PH/69 The HTML version of the specification is now built in a directory called
1334 spec_html instead of spec.html, because the latter looks like a path with
1335 a MIME-type, and this confuses some software.
1337 PH/70 Catch two compiler warnings in sieve.c.
1339 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1340 function verify_get_ident() calls ip_connect() to connect a socket, but
1341 if the "connect()" function timed out, ip_connect() used to close the
1342 socket. However, verify_get_ident() also closes the socket later, and in
1343 between Exim writes to the log, which may get opened at this point. When
1344 the socket was closed in ip_connect(), the log could get the same file
1345 descriptor number as the socket. This naturally causes chaos. The fix is
1346 not to close the socket in ip_connect(); the socket should be closed by
1347 the function that creates it. There was only one place in the code where
1348 this was missing, in the iplookup router, which I don't think anybody now
1349 uses, but I've fixed it anyway.
1351 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1352 well as to direct DNS lookups. Otherwise the handling of names in host
1353 lists is inconsistent and therefore confusing.
1359 PH/01 Two changes to the default runtime configuration:
1361 (1) Move the checks for relay_from_hosts and authenticated clients from
1362 after to before the (commented out) DNS black list checks.
1364 (2) Add control=submission to the relay_from_hosts and authenticated
1365 clients checks, on the grounds that messages accepted by these
1366 statements are most likely to be submissions.
1368 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1370 (1) Generate an error if the third argument for the ${prvs expansion is
1373 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1376 (3) Reset the variables that are obtained from the first argument of
1377 ${prvscheck and used in the second argument before leaving the code,
1378 because their memory is reclaimed, so using them afterwards may do
1381 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1382 one (it's much easier than Tom thought :-).
1384 (5) Because of (4), we can now allow for the use of $prvscheck_result
1385 inside the third argument.
1387 PH/03 For some reason, the default setting of PATH when running a command from
1388 a pipe transport was just "/usr/bin". I have changed it to
1391 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1392 anything to be listed in the output from -bV.
1394 PH/05 When a filter generated an autoreply, the entire To: header line was
1395 quoted in the delivery log line, like this:
1397 => >A.N.Other <ano@some.domain> <original@ddress> ...
1399 This has been changed so that it extracts the operative address. There
1400 may be more than one such address. If so, they are comma-separated, like
1403 => >ano@some.domain,ona@other.domain <original@ddress> ...
1405 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1406 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1407 not being looked up in the rDNS to get a host name, Exim was showing the
1408 IP address twice in Received: lines, even though the IP addresses were
1409 identical. For example:
1411 Received: from [1.2.3.4] (helo=[1.2.3.4])
1413 However, if the real host name was known, it was omitting the HELO data
1414 if it matched the actual IP address. This has been tidied up so that it
1415 doesn't show the same IP address twice.
1417 PH/07 When both +timestamp and +memory debugging was on, the value given by
1418 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1419 called by the debug printing, thereby overwriting the timestamp buffer.
1420 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1422 PH/08 When the original message was included in an autoreply transport, it
1423 always said "this is a copy of the message, including all the headers",
1424 even if body_only or headers_only was set. It now gives an appropriate
1427 PH/09 Applied a patch from the Sieve maintainer which:
1429 o fixes some comments
1430 o adds the (disabled) notify extension core
1431 o adds some debug output for the result of if/elsif tests
1432 o points to the current vacation draft in the documentation
1433 and documents the missing references header update
1437 o fixes a bug in processing the envelope test (when testing
1438 multiple envelope elements, the last element determinted the
1441 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1442 Electronic Mail") by including:
1444 Auto-submitted: auto-generated
1446 in the messages that it generates (bounce messages and others, such as
1447 warnings). In the case of bounce messages for non-SMTP mesages, there was
1448 also a typo: it was using "Auto_submitted" (underscore instead of
1449 hyphen). Since every message generated by Exim is necessarily in response
1450 to another message, thes have all been changed to:
1452 Auto-Submitted: auto-replied
1454 in accordance with these statements in the RFC:
1456 The auto-replied keyword:
1458 - SHOULD be used on messages sent in direct response to another
1459 message by an automatic process,
1461 - MUST NOT be used on manually-generated messages,
1463 - MAY be used on Delivery Status Notifications (DSNs) and Message
1464 Disposition Notifications (MDNs),
1466 - MUST NOT be used on messages generated by automatic or periodic
1467 processes, except for messages which are automatic responses to
1470 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1471 to the default Received: header definition.
1473 PH/12 Added log selector acl_warn_skipped (default on).
1475 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1476 variables because (a) they are in the wrong storage pool and (b) even if
1477 they were copied, it wouldn't work properly because of the caching.
1479 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1480 checking when decoding. Apparently there are clients that generate
1481 overlong encoded strings. Why am I not surprised?
1483 PH/15 If the first argument of "${if match_address" was not empty, but did not
1484 contain an "@" character, Exim crashed. Now it writes a panic log message
1485 and treats the condition as false.
1487 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1489 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1490 extension "envelope-auth". The code is finished and in agreement with
1491 other implementations, but there is no documentation so far and in fact,
1492 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1493 not changing the active code.
1495 Print executed "if" and "elsif" statements when debugging is used. This
1496 helps a great deal to understand what a filter does.
1498 Document more things not specified clearly in RFC3028. I had all this
1499 sorted out, when out of a sudden new issues came to my mind. Oops."
1501 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1504 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1505 canonical form (as documented). However, after a host name lookup from
1506 the IP address, check_host() was doing a simple string comparison with
1507 addresses acquired from the DNS when checking that the found name did
1508 have the original IP as one of its addresses. Since any found IPv6
1509 addresses are likely to be in abbreviated form, the comparison could
1510 fail. Luckily, there already exists a function for doing the comparison
1511 by converting both addresses to binary, so now that is used instead of
1512 the text comparison.
1514 PH/20 There was another similar case to PH/19, when a complete host name was
1515 given in a host list; looking up its IP address could give an abbreviated
1516 form, whereas the current host's name might or might not be abbreviated.
1517 The same fix has been applied.
1523 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1524 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1527 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1528 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1530 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1532 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1533 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1534 Also, for "Auto-Submitted", treat anything other than "no" as
1535 non-personal, in accordance with RFC 3834. (Previously it treated
1536 anything starting "auto-" as non-personal.)
1538 TF/01 The control=submission/name=... option had a problem with syntax
1539 errors if the name included a slash character. The /name= option
1540 now slurps the rest of the string, so it can include any characters
1541 but it must come last in the list of options (after /sender_retain
1544 PH/05 Some modifications to the interface to the fake nameserver for the new
1552 TK/01 Added the "success_on_redirect" address verification option. See
1553 NewStuff for rationale and an example.
1555 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1557 PH/02 Patch to exigrep to allow it to work on syslog lines.
1559 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1560 fread() to skip over the body file's header line, because in Cygwin the
1561 header line is locked and is inaccessible.
1563 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1564 co-exist for some time) to make it clear that it is the Exim ID that is
1565 referenced, not the Message-ID: header line.
1567 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1568 string_format() function, because snprintf() does not exist on all
1571 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1572 ${stat: expansion item.
1574 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1575 protocol synchronization error", to keep the pedants happy.
1577 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1578 well as for IRIX systems, when gcc is being used. See the host.c source
1581 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1583 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1586 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1587 transport and to the smtp transport in LMTP mode.
1589 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1591 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1592 run a filter in a subprocess. This could lead to confusion in subsequent
1593 lookups in the parent process. There should also be a search_tidyup() at
1594 the end of the subprocess.
1596 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1597 only if the host matched helo_try_verify_hosts, which caused the
1598 verification to occur when the EHLO/HELO command was issued. The ACL just
1599 tested the remembered result. Now, if a previous verification attempt has
1600 not happened, "verify = helo" does it there and then.
1602 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1604 TK/03 Fix log output including CR from clamd.
1606 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1607 exist provoked a memory error which could cause a segfault.
1609 PH/15 Installed PCRE 6.2
1611 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1613 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1614 of the problem. Specifically, suggested +O2 rather than +O1 for the
1617 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1619 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1620 record, the port was not correctly logged when the outgoing_port log
1621 selector was set (it logged the transort's default port).
1623 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1624 fallback_hosts, and "hosts" in the smtp transport.
1626 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1627 host errors such as "Connection refused".
1629 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1630 authentication with radiusclient 0.4.9:
1632 - Error returned from rc_read_config was caught wrongly
1633 - Username/password not passed on to radius server due to wrong length.
1635 The presumption is that some radiusclient API changes for 4.51/PH/17
1636 were not taken care of correctly. The code is still untested by me (my
1637 Linux distribution still has 0.3.2 of radiusclient), but it was
1638 contributed by a Radius user.
1640 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1641 expanding the "port" option of the smtp transport.
1643 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1644 while reading a MIME header. Thanks to Tom Hughes for a patch.
1646 PH/24 Include config.h inside local_scan.h so that configuration settings are
1649 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1650 that in an ACL for RCPT (for example), you can examine exactly what was
1653 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1654 commands, but it was not correctly comparing the address with the actual
1655 client host address. Thus, it would show the EHLO address in Received:
1656 header lines when this was not necessary.
1658 PH/27 Added the % operator to ${eval:}.
1660 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1661 it should be ignoring failures (because with -C, for example, it has lost
1662 privilege). It wasn't ignoring creation failures other than "already
1665 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1668 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1669 because an input line was too long, either on its own, or by virtue of
1670 too many continuations, the temporary file was not being removed, and the
1671 return code was incorrect.
1673 PH/31 Missing "BOOL" in function definition in filtertest.c.
1675 PH/32 Applied Sieve patches from the maintainer.
1677 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1679 PH/33 Added "verify = not_blind".
1681 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1682 Local/Makefile (with some defaults set). These are used in built scripts
1683 such as exicyclog, but they have never been used in the exim_install
1684 script (though there are many overriding facilities there). I have
1685 arranged that the exim_install script now takes note of these two
1688 PH/35 Installed configuration files for Dragonfly.
1690 PH/36 When a locally submitted message by a trusted user did not contain a
1691 From: header, and the sender address was obtained from -f or from an SMTP
1692 MAIL command, and the trusted user did not use -F to supply a sender
1693 name, $originator_name was incorrectly used when constructing a From:
1694 header. Furthermore, $originator_name was used for submission mode
1695 messages from external hosts without From: headers in a similar way,
1696 which is clearly wrong.
1698 PH/37 Added control=suppress_local_fixups.
1700 PH/38 When log_selector = +received_sender was set, and the addition of the
1701 sender made the log line's construction buffer exactly full, or one byte
1702 less than full, an overflow happened when the terminating "\n" was
1705 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1706 when the result of a list match is failure because a DNS lookup failed.
1708 PH/40 RM_COMMAND is now used in the building process.
1710 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1711 the "build-* directories that it finds.
1713 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1714 address in a domain literal was a prefix of an interface address.
1716 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1717 when verifying a sender address, unless rewrite_headers is false.
1719 PH/44 Wrote a long comment about why errors_to addresses are verified as
1720 recipients, not senders.
1722 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1723 the ratelimit ACL was added.
1725 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1727 PH/46 Added extra information about PostgreSQL errors to the error string.
1729 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1730 avoiding the need to install special zones in a real server. This is
1731 backwards compatible; if it can't find the fake resolver, it drops back.
1732 Thus, both old and new test suites can be run.
1734 TF/02 Added util/ratelimit.pl
1736 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1737 clock is set back in time.
1739 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1740 Candler <B.Candler@pobox.com>.
1742 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1743 OK for addresses that are the result of redirections.
1745 PH/48 A number of further additions for the benefit of the new test suite,
1746 including a fake gethostbyname() that interfaces to the fake DNS resolver
1749 TF/06 The fix for widen_domains has also been applied to qualify_single and
1750 search_parents which are the other dnslookup options that can cause
1753 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1756 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1758 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1759 service" if it encountered an MX record with an empty target hostname.
1760 The message is now "an MX or SRV record indicated no SMTP service".
1762 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1763 if the DNS of the sending site is misconfigured. This is quite a
1764 common situation. This change restores the behaviour of treating a
1765 helo verification defer as a failure.
1767 PH/51 If self=fail was set on a router, the bounce message did not include the
1768 actual error message.
1774 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1776 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1777 command itself ended in error, the underlying message about the transport
1778 filter timeout was being overwritten with the pipe command error. Now the
1779 underlying error message should be appended to the second error message.
1781 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1783 PH/02 Reduce the amount of output that "make" produces by default. Full output
1784 can still be requested.
1786 PH/03 The warning log line about a condition test deferring for a "warn" verb
1787 was being output only once per connection, rather than after each
1788 occurrence (because it was using the same function as for successful
1789 "warn" verbs). This seems wrong, so I have changed it.
1791 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1792 it should not have, which might have caused a crash in the right
1793 circumstances, but probably never did.
1795 PH/04 Installed a modified version of Tony Finch's patch to make submission
1796 mode fix the return path as well as the Sender: header line, and to
1797 add a /name= option so that you can make the user's friendly name appear
1800 TF/03 Added the control = fakedefer ACL modifier.
1802 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1803 Mark Lowes for thorough testing.
1805 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1807 TK/03 Merged latest SRS patch from Miles Wilton.
1809 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1810 with the definition in sysexits.h (which is #included earlier).
1811 Fortunately, Exim does not actually use EX_OK. The code used to try to
1812 preserve the sysexits.h value, by assumimg that macro definitions were
1813 scanned for macro replacements. I have been disabused of this notion,
1814 so now the code just undefines EX_OK before #including unistd.h.
1816 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1817 in the smtp transport. When a block could not be written in a single
1818 write() function, the timeout was being re-applied to each part-write.
1819 This seems wrong - if the receiver was accepting one byte at a time it
1820 would take for ever. The timeout is now adjusted when this happens. It
1821 doesn't have to be particularly precise.
1823 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1824 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1826 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1827 without a domain if the check to <postmaster@domain> fails.
1829 SC/01 Eximstats: added -xls and the ability to specify output files
1830 (patch written by Frank Heydlauf).
1832 SC/02 Eximstats: use FileHandles for outputing results.
1834 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1836 SC/04 Eximstats: fixed display of large numbers with -nvr option
1838 SC/05 Eximstats: fixed merging of reports with empty tables.
1840 SC/06 Eximstats: added the -include_original_destination flag
1842 SC/07 Eximstats: removed tabs and trailing whitespace.
1844 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1846 TK/06 MBOX spool code: Add real "From " MBOX separator line
1847 so the .eml file is really in mbox format (even though
1848 most programs do not really care). Patch from Alex Miller.
1850 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1851 The latter is generated from $received_to and is only set if the
1852 message has one envelope recipient. SA can use these headers,
1853 obviously out-of-the-box. Patch from Alex Miller.
1855 PH/08 The ${def test on a variable was returning false if the variable's
1856 value was "0", contrary to what the specification has always said!
1857 The result should be true unless the variable is empty.
1859 PH/09 The syntax error of a character other than { following "${if
1860 def:variable_name" (after optional whitespace) was not being diagnosed.
1861 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1862 accidental colon was present, for example, could give incorrect results.
1864 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1865 result is used (not including appendfile, where other changes are about
1868 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1869 This involved changing a lot of size variables from int to off_t. It
1870 should work with maildirs and everything.
1872 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1873 spamd dying while we are connected to it.
1875 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1878 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1879 function for BDB 4.3.
1881 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1883 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1884 the sources for PCRE have been re-arranged and more files are now
1887 PH/15 The code I had for printing potentially long long variables in PH/11
1888 above was not the best (it lost precision). The length of off_t variables
1889 is now inspected at build time, and an appropriate printing format (%ld
1890 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1891 to be "long long int" or "long int". This is needed for the internal
1892 formatting function string_vformat().
1894 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1895 the configuration file to be ":syslog", then the script "guesses" where
1896 the logs files are, rather than using the compiled in default. In our
1897 case the guess is not the same as the compiled default, so the script
1898 suddenly stopped working when I started to use syslog. The patch checks
1899 to see if log_file_path is "". If so, it attempts to read it from exim
1900 with no configuration file to get the compiled in version, before it
1901 falls back to the previous guessing code."
1903 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1904 implementing BATV in an Exim configuration. See NewStuff for the gory
1907 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1908 Makefile that are specific to HP-UX.
1910 PH/18 If the "use_postmaster" option was set for a recipient callout together
1911 with the "random" option, the postmaster address was used as the MAIL
1912 FROM address for the random test, but not for the subsequent recipient
1913 test. It is now used for both.
1915 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1916 patch removes a few documentation additions to RFC 3028, because the
1917 latest draft now contains them. It adds the new en;ascii-case comparator
1918 and a new error check for 8bit text in MIME parts. Comparator and
1919 require names are now matched exactly. I enabled the subaddress
1920 extension, but it is not well tested yet (read: it works for me)."
1922 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1923 rework some of the code of TK/09 above to avoid the hardwired use of
1924 "%lld" and "long long". Replaced the call to snprintf() with a call to
1927 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1928 records point to non-existent hosts", "retry timeout exceeded", and
1929 "retry time not reached for any host after a long failure period".
1931 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1932 experimental DomainKeys support:
1934 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1935 (2) On an error such as an illegally used "control", the wrong name for
1936 the control was given.
1938 These problems did NOT occur unless DomainKeys support was compiled.
1940 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1942 PH/24 Added ${if match_ip condition.
1944 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1945 databases so that it will be absolutely obvious if a crash occurs in the
1946 DB library. This is a regular occurrence (often caused by mis-matched
1949 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1950 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1951 picked up on a user's system that detects such things. There doesn't seem
1952 to be a gcc warning option for this - only an attribute that has to be
1953 put on the function's prototype. It seems that in Fedora Core 4 they have
1954 set this on a number of new functions. No doubt there will be more in due
1957 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1958 specify a transport. However, if an address that was verified by such a
1959 router was the subject of a callout, Exim crashed because it tried to
1960 read the rcpt_include_affixes from the non-existent transport. Now it
1961 just assumes that the setting of that option is false. This bug was
1962 introduced by 4.51/PH/31.
1964 PH/28 Changed -d+all to exclude +memory, because that information is very
1965 rarely of interest, but it makes the output a lot bigger. People tend to
1966 do -d+all out of habit.
1968 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1969 code in os-type was giving problems when libc.so lives in lib64, like on
1972 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1973 aren't the modern standard, and it seems that some systems' include files
1974 don't always have them. Exim was already checking for some of the newer
1975 ones like T_AAAA, and defining it itself. I've added checks for all the
1976 record types that Exim uses.
1978 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1979 not automatically generating a new one, as it is supposed to. This
1980 prevented TLS from working. If the file did exist, but contained invalid
1981 data, a new version was generated, as expected. It was only the case of a
1982 non-existent file that was broken.
1984 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1985 with a change in libdomainkeys > 0.64.
1987 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1988 from DNS. If the selector record carries the flag, it now has
1989 precedence over the domain-wide flag.
1991 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1993 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1994 the use of a port name that isn't defined in /etc/services) occurred, the
1995 message was deferred as in a normal delivery, and thus remained on the
1996 spool, instead of being failed because of the mua_wrapper setting. This
1997 is now fixed, and I tidied up some of the mua_wrapper messages at the
2000 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2001 the messages in a hash of arrays rather than using individual hashes.
2002 This is a bit cleaner and results in dramatic memory savings, albeit
2003 at a slight CPU cost.
2005 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2006 as requested by Marc Sherman.
2008 SC/10 Eximstats: added histograms for user specified patterns as requested
2011 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2013 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2014 fopen() in the content-scanning modules that did not already have it.
2020 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2021 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2023 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2025 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2027 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2028 to test Sieve filters that use "vacation".
2030 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2031 that changes the way the GnuTLS parameters are stored in the cache file.
2032 The new format can be generated externally. For backward compatibility,
2033 if the data in the cache doesn't make sense, Exim assumes it has read an
2034 old-format file, and it generates new data and writes a new file. This
2035 means that you can't go back to an older release without removing the
2038 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2039 work if there are any delivery delays because "one_time" forces the
2040 parent to be marked "delivered", so its unseen clone is never tried
2041 again. For this reason, Exim now forbids the simultaneous setting of
2044 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2045 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2046 entry if you want to know the details. The fix, however, neglected to
2047 consider the case when local delivery batching is involved. The test for
2048 "previously delivered" was not happening when checking to see if an
2049 address could be batched with a previous (undelivered) one; under
2050 certain circumstances this could lead to multiple deliveries to the same
2053 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2054 in its include files, and this causes problems building Exim.
2056 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2057 header_syntax) but Exim was just ignoring anything given after a slash.
2058 In particular, this caused confusion with an attempt to use "verify =
2059 reverse_host_lookup/defer_ok". An error is now given when options are
2060 supplied for verify items that do not have them. (Maybe reverse_host_
2061 lookup should have a defer_ok option, but that's a different point.)
2063 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2064 defined by RFC 821) to 2048, because there were problems with some AUTH
2065 commands, and RFC 1869 says the size should be increased for extended
2066 SMTP commands that take arguments.
2068 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2071 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2072 "unknown" error; now it says that the functionality isn't in the binary.
2074 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2075 an address' error message when a string expansion fails (syntax or
2076 whatever). Otherwise the password may appear in the log. Following change
2077 PH/42 below, there is no longer a chance of it appearing in a bounce
2080 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2082 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2083 of its addresses was ever tried. (Bugzilla bug #2.)
2085 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2086 the result incorrectly in the debug output. (It correctly added a newline
2087 to what was transported.)
2089 TF/01 Added $received_time.
2091 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2092 commented out examples of how to interface to a virus scanner and to
2093 SpamAssassin. Also added commented examples of av_scanner and
2094 spamd_address settings.
2096 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2097 and controls are allowed in which ACLs. There were a couple of minor
2098 errors. Some of the entries in the conditions table (which is a table of
2099 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2100 negation of where the condition IS allowed.
2102 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2104 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2105 header file does not have a version number, so I've had to invent a new
2106 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2107 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2108 radiusclient), but it was contributed by a Radius user.
2110 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2111 files or directories, for OpenSSL.
2113 PH/19 When an Exim process that is running as root has to create an Exim log
2114 file, it does so in a subprocess that runs as exim:exim so as to get the
2115 ownership right at creation (otherwise, other Exim processes might see
2116 the file with the wrong ownership). There was no test for failure of this
2117 fork() call, which would lead to the process getting stuck as it waited
2118 for a non-existent subprocess. Forks do occasionally fail when resources
2119 run out. I reviewed all the other calls to fork(); they all seem to check
2122 PH/20 When checking for unexpected SMTP input at connect time (before writing
2123 the banner), Exim was not dealing correctly with a non-positive return
2124 from the read() function. If the client had disconnected by this time,
2125 the result was a log entry for a synchronization error with an empty
2126 string after "input=" when read() returned zero. If read() returned -1
2127 (an event I could not check), uninitialized data bytes were printed.
2128 There were reports of junk text (parts of files, etc) appearing after
2131 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2133 PH/22 Added support for macro redefinition, and (re)definition in between
2134 driver and ACL definitions.
2136 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2137 forgetting to use the resulting value; it was using the unexpanded value.
2139 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2140 hadn't been configured. The fix is from Juergen Kreileder, who
2141 understands it better than I do:
2143 "Here's what I see happening with three configured cyrus_sasl
2144 authenticators configured (plain, login, cram-md5):
2146 On startup auth_cyrus_sasl_init() gets called for each of these.
2147 This means three calls to sasl_listmech() without a specified mech_list.
2148 => SASL tests which mechs of all available mechs actually work
2149 => three warnings about OTP not working
2150 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2152 With the patch, sasl_listmech() also gets called three times. But now
2153 SASL's mech_list option is set to the server_mech specified in the the
2154 authenticator. Or in other words, the answer from sasl_listmech()
2155 gets limited to just the mech you're testing for (which is different
2157 => the return list contains just 'plain' or 'login', 'cram-md5' or
2158 nothing depending on the value of ob->server_mech.
2160 I've just tested the patch: Authentication still works fine,
2161 unavailable mechs specified in the exim configuration are still
2162 caught, and the auth.log warnings about OTP are gone."
2164 PH/25 When debugging is enabled, the contents of the command line are added
2165 to the debugging output, even when log_selector=+arguments is not
2168 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2169 answer is "GNU", and only if the return is "GNU/something" is the answer
2172 PH/27 $acl_verify_message is now set immediately after the failure of a
2173 verification in an ACL, and so is available in subsequent modifiers. In
2174 particular, the message can be preserved by coding like this:
2176 warn !verify = sender
2177 set acl_m0 = $acl_verify_message
2179 Previously, $acl_verify_message was set only while expanding "message"
2180 and "log_message" when a very denied access.
2182 PH/28 Modified OS/os.c-Linux with
2184 -#ifndef OS_LOAD_AVERAGE
2185 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2187 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2188 nomenclature these days.)
2190 PH/29 Installed patch from the Sieve maintainer that adds the options
2191 sieve_useraddress and sieve_subaddress to the redirect router.
2193 PH/30 In these circumstances:
2194 . Two addresses routed to the same list of hosts;
2195 . First host does not offer TLS;
2196 . First host accepts first address;
2197 . First host gives temporary error to second address;
2198 . Second host offers TLS and a TLS session is established;
2199 . Second host accepts second address.
2200 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2201 and peerdn, if requested) that were in fact used only for the second
2204 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2205 attention to any local part prefix or suffix that was matched by the
2206 router that accepted the address. It now behaves in the same way as it
2207 does for delivery: the affixes are removed from the local part unless
2208 rcpt_include_affixes is set on the transport.
2210 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2211 timeout during the DATA phase of an incoming message.
2213 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2214 applied a patch sanctioned by the Sieve maintainer.
2216 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2217 the uid or gid is negative. A case of a negative gid caused this to be
2218 noticed. The fix allows for either to be negative.
2220 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2221 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2224 PH/36 The change PH/12 above was broken. Fixed it.
2226 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2227 the grounds that routing the same address twice would always produce the
2228 same answer. This might have been true once, but it is certainly no
2229 longer true now. Routing a child address may depend on the previous
2230 routing that produced that child. Some complicated redirection strategies
2231 went wrong when messages had multiple recipients, and made Exim's
2232 behaviour dependent on the order in which the addresses were given.
2234 I have moved the duplicate checking until after the routing is complete.
2235 Exim scans the addresses that are assigned to local and remote
2236 transports, and removes any duplicates. This means that more work will be
2237 done, as duplicates will always all be routed, but duplicates are
2238 presumably rare, so I don't expect this is of any significance.
2240 For deliveries to pipes, files, and autoreplies, the duplicate checking
2241 still happens during the routing process, since they are not going to be
2244 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2245 It corrects a timeout issue with spamd. This is Ian's comment: "The
2246 background is that sometimes spamd either never reads data from a
2247 connection it has accepted, or it never writes response data. The exiscan
2248 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2249 blindly assumes that writes won't block so it may never time out."
2251 PH/39 Allow G after quota size as well as K and M.
2253 PH/40 The value set for $authenticated_id in an authenticator may not contain
2254 binary zeroes or newlines because the value is written to log lines and
2255 to spool files. There was no check on this. Now the value is run through
2256 the string_printing() function so that such characters are converted to
2257 printable escape sequences.
2259 PH/41 $message_linecount is a new variable that contains the total number of
2260 lines in the message. Compare $body_linecount, which is the count for the
2263 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2264 bounce and delay warning messages, except in certain special cases, which
2267 (a) An SMTP error message from a remote host;
2268 (b) A message specified in a :fail: redirection;
2269 (c) A message specified in a "fail" command in a system filter;
2270 (d) A message specified in a FAIL return from the queryprogram router;
2271 (e) A message specified by the cannot_route_message router option.
2273 In these cases only, Exim does include the error details in bounce and
2274 warning messages. There are also a few cases where bland messages such
2275 as "unrouteable address" or "local delivery error" are given.
2277 PH/43 $value is now also set for the "else" part of a ${run expansion.
2279 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2280 being worked on, but at least Exim now implements the latest version to
2283 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2284 process to complete was treated as a delivery failure, a timeout while
2285 writing the message to the pipe was logged, but erroneously treated as a
2286 successful delivery. Such timeouts include transport filter timeouts. For
2287 consistency with the overall process timeout, these timeouts are now
2288 treated as errors, giving rise to delivery failures by default. However,
2289 there is now a new Boolean option for the pipe transport called
2290 timeout_defer, which, if set TRUE, converts the failures into defers for
2291 both kinds of timeout. A transport filter timeout is now identified in
2294 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2295 systems where "make" and "gmake" are different, calling "gmake" at top
2296 level broke things. I've arranged for the value of $(MAKE) to be passed
2297 from the Makefile to this script so that it can call the same version of
2301 A note about Exim versions 4.44 and 4.50
2302 ----------------------------------------
2304 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2305 changes of various kinds. As a consequence, a big documentation update was
2306 needed. This delayed the release for rather longer than seemed good, especially
2307 in the light of a couple of (minor) security issues. Therefore, the changes
2308 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2309 release. So 4.44 and 4.50 are in effect two different branches that both start
2312 I have left the 4.50 change log unchanged; it contains all the changes since
2313 4.43. The change log for 4.44 is below; many of its items are identical to
2314 those for 4.50. This seems to be the most sensible way to preserve the
2315 historical information.
2321 1. Minor wording change to the doc/README.SIEVE file.
2323 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2324 computation of the current number of files was incorrect.
2326 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2327 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2330 4. Give more explanation in the error message when the command for a transport
2331 filter fails to execute.
2333 5. There are several places where Exim runs a non-Exim command in a
2334 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2335 was being done only for the command run by the queryprogram router. It is
2336 now done for all such subprocesses. The other cases are: ${run, transport
2337 filters, and the commands run by the lmtp and pipe transports.
2339 6. Added CONFIGURE_GROUP build-time option.
2341 7. Some older OS have a limit of 256 on the maximum number of file
2342 descriptors. Exim was using setrlimit() to set 1000 as a large value
2343 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2344 systems. I've change it so that if it can't get 1000, it tries for 256.
2346 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2347 was an oversight, and furthermore, ever since the addition of extra
2348 controls (e.g. 4.43/32), the checks on when to allow different forms of
2349 "control" were broken. There should now be diagnostics for all cases when a
2350 control that does not make sense is encountered.
2352 9. Added the /retain_sender option to "control=submission".
2354 10. $recipients is now available in the predata ACL (oversight).
2356 11. Tidy the search cache before the fork to do a delivery from a message
2357 received from the command line. Otherwise the child will trigger a lookup
2358 failure and thereby defer the delivery if it tries to use (for example) a
2359 cached ldap connection that the parent has called unbind on.
2361 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2362 of $address_data from the recipient verification was clobbered by the
2363 sender verification.
2365 13. The value of address_data from a sender verification is now available in
2366 $sender_address_data in subsequent conditions in the ACL statement.
2368 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2370 15. Added a new option "connect=<time>" to callout options, to set a different
2373 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2374 was its contents. (It was OK if the option was not defined at all.)
2376 17. A "Completed" log line is now written for messages that are removed from
2377 the spool by the -Mrm option.
2379 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2380 information about exactly what failed.
2382 19. Added -dd to debug only the daemon process.
2384 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2385 handles timeouts, both on the server side and network timeouts. Renamed the
2386 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2388 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2389 It is now set to "smtps".
2391 22. $host_address is now set to the target address during the checking of
2392 ignore_target_hosts.
2394 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2395 being passed; this would have caused $sender_host_name to have been used if
2396 matching the list had actually called for a host name (not very likely,
2397 since this list is usually IP addresses). A host name is now passed as
2400 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2401 code that specifies a non-restarting handler (typically sigaction() in
2402 modern systems) in an attempt to fix a rare and obscure crash bug.
2404 25. Narrowed the window for a race in the daemon that could cause it to ignore
2405 SIGCHLD signals. This is not a major problem, because they are used only to
2406 wake it up if nothing else does.
2408 26. A malformed maildirsize file could cause Exim to calculate negative values
2409 for the mailbox size or file count. Odd effects could occur as a result.
2410 The maildirsize information is now recalculated if the size or filecount
2413 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2414 support for a long time. Removed HAVE_SYS_VFS_H.
2416 28. Installed the latest version of exipick from John Jetmore.
2418 29. In an address list, if the pattern was not a regular expression, an empty
2419 subject address (from a bounce message) matched only if the pattern was an
2420 empty string. Non-empty patterns were not even tested. This was the wrong
2421 because it is perfectly reasonable to use an empty address as part of a
2422 database query. An empty address is now tested by patterns that are
2423 lookups. However, all the other forms of pattern expect the subject to
2424 contain a local part and a domain, and therefore, for them, an empty
2425 address still always fails if the pattern is not itself empty.
2427 30. Exim went into a mad DNS loop when attempting to do a callout where the
2428 host was specified on an smtp transport, and looking it up yielded more
2429 than one IP address.
2431 31. Re-factored the code for checking spool and log partition space into a
2432 function that finds that data and another that does the check. The former
2433 is then used to implement four new variables: $spool_space, $log_space,
2434 $spool_inodes, and $log_inodes.
2436 32. The RFC2047 encoding function was originally intended for short strings
2437 such as real names; it was not keeping to the 75-character limit for
2438 encoded words that the RFC imposes. It now respects the limit, and
2439 generates multiple encoded words if necessary. To be on the safe side, I
2440 have increased the buffer size for the ${rfc2047: expansion operator from
2443 33. It is now permitted to omit both strings after an "if" condition; if the
2444 condition is true, the result is "true". As before, when the second string
2445 is omitted, a false condition yields an empty string. This makes it less
2446 cumbersome to write custom ACL and router conditions.
2448 34. Failure to deliver a bounce message always caused it to be frozen, even if
2449 there was an errors_to setting on the router. The errors_to setting is now
2452 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2453 canonical form (fully expanded) before being placed in
2454 $sender_host_address.
2456 36. The table in the code that translates DNS record types into text (T_A to
2457 "A" for instance) was missing entries for NS and CNAME. It is just possible
2458 that this could have caused confusion if both these types were looked up
2459 for the same domain, because the text type is used as part of Exim's
2460 per-process caching. But the chance of anyone hitting this buglet seems
2463 37. The dnsdb lookup has been extended in a number of ways.
2465 (1) There is a new type, "zns", which walks up the domain tree until it
2466 finds some nameserver records. It should be used with care.
2468 (2) There is a new type, "mxh", which is like "mx" except that it returns
2469 just the host names, not the priorities.
2471 (3) It is now possible to give a list of domains (or IP addresses) to be
2472 looked up. The behaviour when one of the lookups defers can be
2473 controlled by a keyword.
2475 (4) It is now possible to specify the separator character for use when
2476 multiple records are returned.
2478 38. The dnslists ACL condition has been extended: it is now possible to supply
2479 a list of IP addresses and/or domains to be looked up in a particular DNS
2482 39. Added log_selector=+queue_time_overall.
2484 40. When running the queue in the test harness, wait just a tad after forking a
2485 delivery process, to get repeatability of debugging output.
2487 41. Include certificate and key file names in error message when GnuTLS fails
2488 to set them up, because the GnuTLS error message doesn't include the name
2489 of the failing file when there is a problem reading it.
2491 42. Allow both -bf and -bF in the same test run.
2493 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2495 44. The "Exiscan patch" is now merged into the mainline Exim source.
2497 45. Sometimes the final signoff response after QUIT could fail to get
2498 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2499 < 0 before doing a fflush(). This bug looks as though it goes back to the
2500 introduction of TLS in release 3.20, but "sometimes" must have been rare
2501 because the tests only now provoked it.
2503 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2504 (this can affect the format of dates).
2506 47. exim_tidydb, when checking for the continued existence of a message for
2507 which it has found a message-specific retry record, was not finding
2508 messages that were in split spool directories. Consequently, it was
2509 deleting retry records that should have stayed in existence.
2511 48. Steve fixed some bugs in eximstats.
2513 49. The SPA authentication driver was not abandoning authentication and moving
2514 on to the next authenticator when an expansion was forced to fail,
2515 contradicting the general specification for all authenticators. Instead it
2516 was generating a temporary error. It now behaves as specified.
2518 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2519 (the order specifies the preference for clients). The order is now AES256,
2520 AES128, 3DES, ARCFOUR128.
2522 51. Small patch to Sieve code - explicitly set From: when generating an
2525 52. Exim crashed if a remote delivery caused a very long error message to be
2526 recorded - for instance if somebody sent an entire SpamAssassin report back
2527 as a large number of 550 error lines. This bug was coincidentally fixed by
2528 increasing the size of one of Exim's internal buffers (big_buffer) that
2529 happened as part of the Exiscan merge. However, to be on the safe side, I
2530 have made the code more robust (and fixed the comments that describe what
2533 53. Now that there can be additional text after "Completed" in log lines (if
2534 the queue_time_overall log selector is set), a one-byte patch to exigrep
2535 was needed to allow it to recognize "Completed" as not the last thing in
2538 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2539 patch that reportedly fixes this has been added. I am not expert enough to
2540 create a test for it. This is what the patch creator wrote:
2542 "I found a little strange behaviour of ldap code when working with
2543 Windows 2003 AD Domain, where users was placed in more than one
2544 Organization Units. When I tried to give exim partial DN, the exit code
2545 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2546 But simultaneously result of request was absolutely normal ldap result,
2547 so I produce this patch..."
2549 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2550 REFERENCE, so I have modified the code to exclude the patch when that macro
2553 55. Some experimental protocols are using DNS PTR records for new purposes. The
2554 keys for these records are domain names, not reversed IP addresses. The
2555 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2556 leaves it alone. Component reversal etc. now happens only for IP addresses.
2559 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2561 57. Double the size of the debug message buffer (to 2048) so that more of very
2562 long debug lines gets shown.
2564 58. The exicyclog utility now does better if the number of log files to keep
2565 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2567 59. Two changes related to the smtp_active_hostname option:
2569 (1) $smtp_active_hostname is now available as a variable.
2570 (2) The default for smtp_banner uses $smtp_active_hostname instead
2571 of $primary_hostname.
2573 60. The host_aton() function is supposed to be passed a string that is known
2574 to be a valid IP address. However, in the case of IPv6 addresses, it was
2575 not checking this. This is a hostage to fortune. Exim now panics and dies
2576 if the condition is not met. A case was found where this could be provoked
2577 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2578 components; fortuitously, this particular loophole had already been fixed
2579 by change 4.50/55 above.
2581 If there are any other similar loopholes, the new check in host_aton()
2582 itself should stop them being exploited. The report I received stated that
2583 data on the command line could provoke the exploit when Exim was running as
2584 exim, but did not say which command line option was involved. All I could
2585 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2586 running as the user.
2589 61. There was a buffer overflow vulnerability in the SPA authentication code
2590 (which came originally from the Samba project). I have added a test to the
2591 spa_base64_to_bits() function which I hope fixes it.
2594 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2595 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2597 63. The daemon start-up calls getloadavg() while still root for those OS that
2598 need the first call to be done as root, but it missed one case: when
2599 deliver_queue_load_max is set with deliver_drop_privilege. This is
2600 necessary for the benefit of the queue runner, because there is no re-exec
2601 when deliver_drop_privilege is set.
2603 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2604 This has been fixed.
2606 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2607 was in use, was not putting the data itself into the right store pool;
2608 consequently, it could be overwritten for a subsequent message in the same
2609 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2612 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2614 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2615 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2616 the function were treating the return as a boolean value, which happened to
2617 work because 0=false and not-0=true, but is not correct code.
2619 68. The host_aton() function was not handling scoped IPv6 addresses (those
2620 with, for example, "%eth0" on the end) correctly.
2622 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2623 negated items (that is, ~something) in unsigned ints. Some compilers
2624 apparently mutter when there is no cast.
2626 70. If an address verification called from an ACL failed, and did not produce a
2627 user-specific message (i.e. there was only a "system" message), nothing was
2628 put in $acl_verify_message. In this situation, it now puts the system
2631 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2632 session; change 4.31/43 added the unwanted input to the log line - except
2633 that it did not do this in the start of session case. It now does.
2635 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2636 This is wrong and can cause the other end to generate a synchronization
2637 error if it is another Exim or anything else that does the synchronization
2638 check. A QUIT command is no longer sent after a timeout.
2640 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2641 during host lookups.
2643 74. The defer_ok option of callout verification was not working if it was used
2644 when verifying addresses in header lines, that is, for this case:
2646 verify = header_sender/callout=defer_ok
2648 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2649 those file descriptors could be used for SMTP connections. If anything
2650 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2651 could be sent to the SMTP client, causing chaos. The daemon now opens
2652 stdin, stdout, and stderr to /dev/null when it puts itself into the
2655 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2656 log by default. The user can override this with suitable Perl magic.
2658 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2659 the log message when discard triggers, was not working for the DATA ACL or
2660 for the non-SMTP ACL.
2662 78. Error message wording change in sieve.c.
2664 79. If smtp_accept_max_per_host was set, the number of connections could be
2665 restricted to fewer than expected, because the daemon was trying to set up
2666 a new connection before checking whether the processes handling previous
2667 connections had finished. The check for completed processes is now done
2668 earlier. On busy systems, this bug wouldn't be noticed because something
2669 else would have woken the daemon, and it would have reaped the completed
2672 80. If a message was submitted locally by a user whose login name contained one
2673 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2674 It caused a spool format error. I have fixed the spool reading code. A
2675 related problem was that the "from" clause in the Received: line became
2676 illegal because of the space(s). It is now covered by ${quote_local_part.
2678 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2681 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2683 83. Patch from Sieve maintainer to fix unterminated string problem in
2684 "vacation" handling.
2686 84. Some minor changes to the Linux configuration files to help with other
2687 OS variants using glibc.
2689 85. One more patch for Sieve to update vacation handling to latest spec.
2692 ----------------------------------------------------
2693 See the note above about the 4.44 and 4.50 releases.
2694 ----------------------------------------------------
2700 1. Change 4.43/35 introduced a bug that caused file counts to be
2701 incorrectly computed when quota_filecount was set in an appendfile
2704 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2705 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2708 3. Give more explanation in the error message when the command for a transport
2709 filter fails to execute.
2711 4. There are several places where Exim runs a non-Exim command in a
2712 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2713 was being done only for the command run by the queryprogram router. It is
2714 now done for all such subprocesses. The other cases are: ${run, transport
2715 filters, and the commands run by the lmtp and pipe transports.
2717 5. Some older OS have a limit of 256 on the maximum number of file
2718 descriptors. Exim was using setrlimit() to set 1000 as a large value
2719 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2720 systems. I've change it so that if it can't get 1000, it tries for 256.
2722 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2723 was an oversight, and furthermore, ever since the addition of extra
2724 controls (e.g. 4.43/32), the checks on when to allow different forms of
2725 "control" were broken. There should now be diagnostics for all cases when a
2726 control that does not make sense is encountered.
2728 7. $recipients is now available in the predata ACL (oversight).
2730 8. Tidy the search cache before the fork to do a delivery from a message
2731 received from the command line. Otherwise the child will trigger a lookup
2732 failure and thereby defer the delivery if it tries to use (for example) a
2733 cached ldap connection that the parent has called unbind on.
2735 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2736 of $address_data from the recipient verification was clobbered by the
2737 sender verification.
2739 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2740 was its contents. (It was OK if the option was not defined at all.)
2742 11. A "Completed" log line is now written for messages that are removed from
2743 the spool by the -Mrm option.
2745 12. $host_address is now set to the target address during the checking of
2746 ignore_target_hosts.
2748 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2749 being passed; this would have caused $sender_host_name to have been used if
2750 matching the list had actually called for a host name (not very likely,
2751 since this list is usually IP addresses). A host name is now passed as
2754 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2755 code that specifies a non-restarting handler (typically sigaction() in
2756 modern systems) in an attempt to fix a rare and obscure crash bug.
2758 15. Narrowed the window for a race in the daemon that could cause it to ignore
2759 SIGCHLD signals. This is not a major problem, because they are used only to
2760 wake it up if nothing else does.
2762 16. A malformed maildirsize file could cause Exim to calculate negative values
2763 for the mailbox size or file count. Odd effects could occur as a result.
2764 The maildirsize information is now recalculated if the size or filecount
2767 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2768 support for a long time. Removed HAVE_SYS_VFS_H.
2770 18. Updated exipick to current release from John Jetmore.
2772 19. Allow an empty sender to be matched against a lookup in an address list.
2773 Previously the only cases considered were a regular expression, or an
2776 20. Exim went into a mad DNS lookup loop when doing a callout where the
2777 host was specified on the transport, if the DNS lookup yielded more than
2780 21. The RFC2047 encoding function was originally intended for short strings
2781 such as real names; it was not keeping to the 75-character limit for
2782 encoded words that the RFC imposes. It now respects the limit, and
2783 generates multiple encoded words if necessary. To be on the safe side, I
2784 have increased the buffer size for the ${rfc2047: expansion operator from
2787 22. Failure to deliver a bounce message always caused it to be frozen, even if
2788 there was an errors_to setting on the router. The errors_to setting is now
2791 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2792 canonical form (fully expanded) before being placed in
2793 $sender_host_address.
2795 24. Updated eximstats to version 1.33
2797 25. Include certificate and key file names in error message when GnuTLS fails
2798 to set them up, because the GnuTLS error message doesn't include the name
2799 of the failing file when there is a problem reading it.
2801 26. Expand error message when OpenSSL has problems setting up cert/key files.
2804 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2805 (this can affect the format of dates).
2807 28. exim_tidydb, when checking for the continued existence of a message for
2808 which it has found a message-specific retry record, was not finding
2809 messages that were in split spool directories. Consequently, it was
2810 deleting retry records that should have stayed in existence.
2812 29. eximstats updated to version 1.35
2813 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2814 1.35 - bugfix such that pie charts by volume are generated correctly
2816 30. The SPA authentication driver was not abandoning authentication and moving
2817 on to the next authenticator when an expansion was forced to fail,
2818 contradicting the general specification for all authenticators. Instead it
2819 was generating a temporary error. It now behaves as specified.
2821 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2822 (the order specifies the preference for clients). The order is now AES256,
2823 AES128, 3DES, ARCFOUR128.
2825 31. Small patch to Sieve code - explicitly set From: when generating an
2828 32. Exim crashed if a remote delivery caused a very long error message to be
2829 recorded - for instance if somebody sent an entire SpamAssassin report back
2830 as a large number of 550 error lines. This bug was coincidentally fixed by
2831 increasing the size of one of Exim's internal buffers (big_buffer) that
2832 happened as part of the Exiscan merge. However, to be on the safe side, I
2833 have made the code more robust (and fixed the comments that describe what
2836 33. Some experimental protocols are using DNS PTR records for new purposes. The
2837 keys for these records are domain names, not reversed IP addresses. The
2838 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2839 leaves it alone. Component reversal etc. now happens only for IP addresses.
2842 34. The host_aton() function is supposed to be passed a string that is known
2843 to be a valid IP address. However, in the case of IPv6 addresses, it was
2844 not checking this. This is a hostage to fortune. Exim now panics and dies
2845 if the condition is not met. A case was found where this could be provoked
2846 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2847 components; fortuitously, this particular loophole had already been fixed
2848 by change 4.50/55 or 4.44/33 above.
2850 If there are any other similar loopholes, the new check in host_aton()
2851 itself should stop them being exploited. The report I received stated that
2852 data on the command line could provoke the exploit when Exim was running as
2853 exim, but did not say which command line option was involved. All I could
2854 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2855 running as the user.
2858 35. There was a buffer overflow vulnerability in the SPA authentication code
2859 (which came originally from the Samba project). I have added a test to the
2860 spa_base64_to_bits() function which I hope fixes it.
2863 36. The daemon start-up calls getloadavg() while still root for those OS that
2864 need the first call to be done as root, but it missed one case: when
2865 deliver_queue_load_max is set with deliver_drop_privilege. This is
2866 necessary for the benefit of the queue runner, because there is no re-exec
2867 when deliver_drop_privilege is set.
2869 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2870 was in use, was not putting the data itself into the right store pool;
2871 consequently, it could be overwritten for a subsequent message in the same
2872 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2875 38. Sometimes the final signoff response after QUIT could fail to get
2876 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2877 < 0 before doing a fflush(). This bug looks as though it goes back to the
2878 introduction of TLS in release 3.20, but "sometimes" must have been rare
2879 because the tests only now provoked it.
2885 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2886 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2887 Now it returns an integer. A number of calls were still expecting a T/F
2888 return. Fortuitously, in all cases, the tests worked in OK situations,
2889 which is the norm. However, things would have gone wrong on any write
2890 failures on the smtp file descriptor. This function is used when sending
2891 messages over SMTP and also when doing verify callouts.
2893 2. When Exim is called to do synchronous delivery of a locally submitted
2894 message (the -odf or -odi options), it no longer closes stderr before doing
2897 3. Implemented the mua_wrapper option.
2899 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2901 5. Implemented the functions header_remove(), header_testname(),
2902 header_add_at_position(), and receive_remove_recipient(), and exported them
2905 6. If an ACL "warn" statement specified the addition of headers, Exim already
2906 inserted X-ACL-Warn: at the start if there was no header name. However, it
2907 was not making this test for the second and subsequent header lines if
2908 there were newlines in the string. This meant that an invalid header could
2909 be inserted if Exim was badly configured.
2911 7. Allow an ACL "warn" statement to add header lines at the start or after all
2912 the Received: headers, as well as at the end.
2914 8. Added the rcpt_4xx retry error code.
2916 9. Added postmaster_mailfrom=xxx to callout verification option.
2918 10. Added mailfrom=xxxx to the callout verification option, for verify=
2921 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2922 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2924 12. Inserted some casts to stop certain compilers warning when using pointer
2925 differences as field lengths or precisions in printf-type calls (mostly
2926 affecting debugging statements).
2928 13. Added optional readline() support for -be (dynamically loaded).
2930 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2931 same clock tick as a message's arrival, so that its received time was the
2932 same as the "first fail" time on the retry record, and that message
2933 remained on the queue past the ultimate address timeout, every queue runner
2934 would try a delivery (because it was past the ultimate address timeout) but
2935 after another failure, the ultimate address timeout, which should have then
2936 bounced the address, did not kick in. This was a "< instead of <=" error;
2937 in most cases the first failure would have been in the next clock tick
2938 after the received time, and all would be well.
2940 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2941 being recognized when the domain list was tested by the match_domain
2942 condition in an expansion string.
2944 16. Added the ${str2b64: operator.
2946 17. Exim was always calling setrlimit() to set a large limit for the number of
2947 processes, without checking whether the existing limit was already
2948 adequate. (It did check for the limit on file descriptors.) Furthermore,
2949 errors from getrlimit() and setrlimit() were being ignored. Now they are
2950 logged to the main and panic logs, but Exim does carry on, to try to do its
2951 job under whatever limits there are.
2953 18. Imported PCRE 5.0.
2955 19. Trivial typo in log message " temporarily refused connection" (the leading
2958 20. If the log selector return_path_on_delivery was set and an address was
2959 redirected to /dev/null, the delivery process crashed because it assumed
2960 that a return path would always be set for a "successful" delivery. In this
2961 case, the whole delivery is bypassed as an optimization, and therefore no
2964 21. Internal re-arrangement: the function for sending a challenge and reading
2965 a response while authentication was assuming a zero-terminated challenge
2966 string. It's now changed to take a pointer and a length, to allow for
2967 binary data in such strings.
2969 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2971 23. Exim was not respecting finduser_retries when seeking the login of the
2972 uid under which it was called; it was always trying 10 times. (The default
2973 setting of finduser_retries is zero.) Also, it was sleeping after the final
2974 failure, which is pointless.
2976 24. Implemented tls_on_connect_ports.
2978 25. Implemented acl_smtp_predata.
2980 26. If the domain in control=submission is set empty, Exim assumes that the
2981 authenticated id is a complete email address when it generates From: or
2982 Sender: header lines.
2984 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2985 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2986 chown and chgrp in /bin and hostname in /usr/bin.
2988 28. Exim was keeping the "process log" file open after each use, just as it
2989 does for the main log. This opens the possibility of it remaining open for
2990 long periods when the USR1 signal hits a daemon. Occasional processlog
2991 errors were reported, that could have been caused by this. Anyway, it seems
2992 much more sensible not to leave this file open at all, so that is what now
2995 29. The long-running daemon process does not normally write to the log once it
2996 has entered its main loop, and it closes the log before doing so. This is
2997 so that log files can straightforwardly be renamed and moved. However,
2998 there are a couple of unusual error situations where the daemon does write
2999 log entries, and I had neglected to close the log afterwards.
3001 30. The text of an SMTP error response that was received during a remote
3002 delivery was being truncated at 512 bytes. This is too short for some of
3003 the long messages that one sometimes sees. I've increased the limit to
3006 31. It is now possible to make retry rules that apply only when a message has a
3007 specific sender, in particular, an empty sender.
3009 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3010 it possible to be selective about when SMTP synchronization is enforced.
3012 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3014 32. Implemented hosts_connection_nolog.
3016 33. Added an ACL for QUIT.
3018 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3021 35. Added mailbox_size and mailbox_filecount to appendfile.
3023 36. Added control = no_multiline_responses to ACLs.
3025 37. There was a bug in the logic of the code that waits for the clock to tick
3026 in the case where the clock went backwards by a substantial amount such
3027 that the microsecond fraction of "now" was more than the microsecond
3028 fraction of "then" (but the whole seconds number was less).
3030 38. Added support for the libradius Radius client library this is found on
3031 FreeBSD (previously only the radiusclient library was supported).
3037 1. When certain lookups returned multiple values in the form name=value, the
3038 quoting of the values was not always being done properly. Specifically:
3039 (a) If the value started with a double quote, but contained no whitespace,
3041 (b) If the value contained whitespace other than a space character (i.e.
3042 tabs or newlines or carriage returns) it was not quoted.
3043 This fix has been applied to the mysql and pgsql lookups by writing a
3044 separate quoting function and calling it from the lookup code. The fix
3045 should probably also be applied to nisplus, ibase and oracle lookups, but
3046 since I cannot test any of those, I have not disturbed their existing code.
3048 2. A hit in the callout cache for a specific address caused a log line with no
3049 reason for rejecting RCPT. Now it says "Previous (cached) callout
3050 verification failure".
3052 3. There was an off-by-one bug in the queryprogram router. An over-long
3053 return line was truncated at 256 instead of 255 characters, thereby
3054 overflowing its buffer with the terminating zero. As well as fixing this, I
3055 have increased the buffer size to 1024 (and made a note to document this).
3057 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3058 when Exim is waiting for an SMTP response from a remote server, Exim
3059 restarts its select() call on the socket, thereby resetting its timeout.
3060 This is not a problem when such interrupts are rare. Somebody set up a cron
3061 job to run exiwhat every 2 minutes, which is less than the normal select()
3062 timeout (5 or 10 minutes). This meant that the select() timeout never
3063 kicked in because it was always reset. I have fixed this by comparing the
3064 time when an interrupt arrives with the time at the start of the first call
3065 to select(). If more time than the timeout has elapsed, the interrupt is
3066 treated as a timeout.
3068 5. Some internal re-factoring in preparation for the addition of Sieve
3069 extensions (by MH). In particular, the "personal" test is moved to a
3070 separate function, and given an option for scanning Cc: and Bcc: (which is
3071 not set for Exim filters).
3073 6. When Exim created an email address using the login of the caller as the
3074 local part (e.g. when creating a From: or Sender: header line), it was not
3075 quoting the local part when it contained special characters such as @.
3077 7. Installed new OpenBSD configuration files.
3079 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3080 try to make them clearer.
3082 9. Callout options, other than the timeout value, were being ignored when
3083 verifying sender addresses in header lines. For example, when using
3085 verify = header_sender/callout=no_cache
3087 the cache was (incorrectly) being used.
3089 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3090 only the Cygwin environment.
3092 11. When return_path_on_delivery was set as a log selector, if different remote
3093 addresses in the same message used different return paths and parallel
3094 remote delivery occurred, the wrong values would sometimes be logged.
3095 (Whenever a remote delivery process finished, the return path value from
3096 the most recently started remote delivery process was logged.)
3098 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3099 lines when AUTH and/or TLS are in use. This is the "received protocol"
3100 field. Exim used to use "asmtp" for authenticated SMTP, without any
3101 indication (in the protocol name) for TLS use. Now it follows the RFC and
3102 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3103 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3104 names appear in log lines as well as in Received: header lines.
3106 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3107 extensions, and comparison tests, and to fix some bugs.
3109 14. Changes to the "personal" filter test:
3111 (1) The test was buggy in that it was just doing the equivalent of
3112 "contains" tests on header lines. For example, if a user's address was
3113 anne@some.where, the "personal" test would incorrectly be true for
3115 To: susanne@some.where
3117 This test is now done by extracting each address from the header in turn,
3118 and checking the entire address. Other tests that are part of "personal"
3119 are now done using regular expressions (for example, to check local parts
3120 of addresses in From: header lines).
3122 (2) The list of non-personal local parts in From: addresses has been
3123 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3124 taken from the Sieve specification recommendations.
3126 (3) If the message contains any header line starting with "List-" it is
3127 treated as non-personal.
3129 (4) The test for "circular" in the Subject: header line has been removed
3130 because it now seems ill-conceived.
3132 15. Minor typos in src/EDITME comments corrected.
3134 16. Installed latest exipick from John Jetmore.
3136 17. If headers_add on a router specified a text string that was too long for
3137 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3138 of string_sprintf() is now avoided.
3140 18. $message_body_size was not set (it was always zero) when running the DATA
3141 ACL and the local_scan() function.
3143 19. For the "mail" command in an Exim filter, no default was being set for
3144 the once_repeat time, causing a random time value to be used if "once" was
3145 specified. (If the value happened to be <= 0, no repeat happened.) The
3146 default is now 0s, meaning "never repeat". The "vacation" command was OK
3147 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3148 (I found it when inspecting the code).
3150 20. There is now an overall timeout for performing a callout verification. It
3151 defaults to 4 times the callout timeout, which applies to individual SMTP
3152 commands during the callout. The overall timeout applies when there is more
3153 than one host that can be tried. The timeout is checked before trying the
3154 next host. This prevents very long delays if there are a large number of
3155 hosts and all are timing out (e.g. when the network connections are timing
3156 out). The value of the overall timeout can be changed by specifying an
3157 additional sub-option for "callout", called "maxwait". For example:
3159 verify = sender/callout=5s,maxwait=20s
3161 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3162 to the end before writing, but this should make it even safer).
3164 22. Exim was forgetting that it had advertised PIPELINING for the second and
3165 subsequent messages on an SMTP connection. It was also not resetting its
3166 memory on STARTTLS and an internal HELO.
3168 23. When Exim logs an SMTP synchronization error within a session, it now
3169 records whether PIPELINING has been advertised or not.
3171 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3172 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3173 rather than long int.
3175 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3177 26. Added the never_mail option to autoreply.
3183 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3184 crash if the getsockname() call failed; this can happen if a connection is
3185 closed very soon after it is established. The problem was simply in the
3186 order in which certain operations were done, causing Exim to try to write
3187 to the SMTP stream before it had set up the file descriptor. The bug has
3188 been fixed by making things happen in the correct order.
3194 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3195 before the connection was closed, thus losing the rejection response.
3197 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3198 some early Solaris releases, but causes trouble in current releases where
3199 socklen_t is defined.
3201 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3204 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3207 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3208 that releasing the top of it at the end releases what was used for sub-
3209 expansions (unless the block got too big). However, discard this block if
3210 the first thing is a variable or header, so that we can use its block when
3211 it is dynamic (useful for very large $message_headers, for example).
3213 6. Lookups now cache *every* query, not just the most recent. A new, separate
3214 store pool is used for this. It can be recovered when all lookup caches are
3215 flushed. Lookups now release memory at the end of their result strings.
3216 This has involved some general refactoring of the lookup sources.
3218 7. Some code has been added to the store_xxx() functions to reduce the amount
3219 of flapping under certain conditions.
3221 8. log_incoming_interface used to affect only the <= reception log lines. Now
3222 it causes the local interface and port to be added to several more SMTP log
3223 lines, for example "SMTP connection from", and rejection lines.
3225 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3227 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3229 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3230 could be overwritten at the end of the current message (or the start of a
3231 new message if it was set in a HELO ACL). The value is now preserved for
3232 the duration of the SMTP connection.
3234 12. If a transport had a headers_rewrite setting, and a matching header line
3235 contained an unqualified address, that address was qualified, even if it
3236 did not match any rewriting rules. The underlying bug was that the values
3237 of the flags that permit the existence of unqualified sender and recipient
3238 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3239 non-local messages, and by -bnq for local messages) were not being
3240 preserved with the message after it was received.
3242 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3243 "next input=" as part of the text comprising the host identity instead of
3244 the correct text. The code was using the same buffer for two different
3245 strings. However, depending on which order the printing function evaluated
3246 its arguments, the bug did not always show up. Under Linux, for example, my
3247 test suite worked just fine.
3249 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3250 doesn't work with some older versions of Perl. It has been changed to "my",
3251 which in any case is probably the better facility to use.
3253 15. A really picky compiler found some instances of statements for creating
3254 error messages that either had too many or two few arguments for the format
3257 16. The size of the buffer for calls to the DNS resolver has been increased
3258 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3259 for addresses that have a lot of PTR records. This alleviates a problem; it
3260 does not fully solve it.
3262 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3263 buffer now truncates the list and logs the incident, which is the same
3264 action as happens when Exim is looking up a host name and its aliases.
3265 Previously in this situation something unpredictable would happen;
3266 sometimes it was "internal error: store_reset failed".
3268 18. If a server dropped the connection unexpectedly when an Exim client was
3269 using GnuTLS and trying to read a response, the client delivery process
3270 crashed while trying to generate an error log message.
3272 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3273 string, for example:
3275 warn message = H1: something\nH2: something
3277 the text was added as a single header line from Exim's point of view
3278 though it ended up OK in the delivered message. However, searching for the
3279 second and subsequent header lines using $h_h2: did not work. This has been
3280 fixed. Similarly, if a system filter added multiple headers in this way,
3281 the routers could not see them.
3283 20. Expanded the error message when iplsearch is called with an invalid key to
3284 suggest using net-iplsearch in a host list.
3286 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3287 ACLs are no longer actually imposed (and a message to that effect is
3290 22. If a "gecos" field in a passwd entry contained escaped characters, in
3291 particular, if it contained a \" sequence, Exim got it wrong when building
3292 a From: or a Sender: header from that name. A second bug also caused
3293 incorrect handling when an unquoted " was present following a character
3294 that needed quoting.
3296 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3297 was not being matched caselessly.
3299 24. Arranged for all hyphens in the exim.8 source to be escaped with
3302 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3303 properly. Recipient callout cache records were still being keyed to include
3304 the sender, even when use_sender was set false. This led to far more
3305 callouts that were necessary. The sender is no longer included in the key
3306 when use_sender is false.
3308 26. Added "control = submission" modifier to ACLs.
3310 27. Added the ${base62d: operator to decode base 62 numbers.
3312 28. dnsdb lookups can now access SRV records.
3314 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3315 the configuration file.
3317 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3318 (-v) mode. This makes the output for a verbose queue run more intelligible.
3320 31. Added a use_postmaster feature to recipient callouts.
3322 32. Added the $body_zerocount variable, containing the number of binary zero
3323 bytes in the message body.
3325 33. The time of last modification of the "new" subdirectory is now used as the
3326 "mailbox time last read" when there is a quota error for a maildir
3329 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3331 35. Added +ignore_unknown as a special item in host lists.
3333 36. Code for decoding IPv6 addresses in host lists is now included, even if
3334 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3335 address was recognized as an IP address, but was then not correctly decoded
3336 into binary, causing unexpected and incorrect effects when compared with
3343 1. Very minor rewording of debugging text in manualroute to say "list of
3344 hosts" instead of "hostlist".
3346 2. If verify=header_syntax was set, and a header line with an unqualified
3347 address (no domain) and a large number of spaces between the end of the
3348 name and the colon was received, the reception process suffered a buffer
3349 overflow, and (when I tested it) crashed. This was caused by some obsolete
3350 code that should have been removed. The fix is to remove it!
3352 3. When running in the test harness, delay a bit after writing a bounce
3353 message to get a bit more predictability in the log output.
3355 4. Added a call to search_tidyup() just before forking a reception process. In
3356 theory, someone could use a lookup in the expansion of smtp_accept_max_
3357 per_host which, without the tidyup, could leave open a database connection.
3359 5. Added the variables $recipient_data and $sender_data which get set from a
3360 lookup success in an ACL "recipients" or "senders" condition, or a router
3361 "senders" option, similar to $domain_data and $local_part_data.
3363 6. Moved the writing of debug_print from before to after the "senders" test
3366 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3367 problems for message scanning, either using a data ACL, or using
3368 local_scan() because the Received: header was not generated till after they
3369 were called (in order to set the time as the time of reception completion).
3370 I have revised the way this works. The header is now generated after the
3371 body is received, but before the ACL or local_scan() are called. After they
3372 are run, the timestamp in the header is updated.
3378 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3379 before starting a queue runner without re-exec. This happened only when
3380 deliver_drop_privilege was set or when the Exim user was set to root. The
3381 effect of the bug was that timeouts during subsequent deliveries caused
3382 crashes instead of being properly handled. The handler is now left at its
3383 default (and expected) setting.
3385 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3386 message, again when deliver_drop_privilege is set or Exim is run as root.
3387 The bug described in (1) was not present in this case, but the tidying up
3388 of the other signals was missing. I have made the two cases consistent.
3390 3. The ignore_target_hosts setting on a manualroute router was being ignored
3391 for hosts that were looked up using the /MX notation.
3393 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3396 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3397 operated on the sender address. After changing the $sender_address to <>
3398 for the sender address verify, Exim was re-instated it as the original
3399 (before rewriting) address, but remembering that it had rewritten it, so it
3400 wasn't rewriting it again. This bug also had the effect of breaking the
3401 sender address verification caching when the sender address was rewritten.
3403 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3404 This has been changed so that if the ip literal address matches
3405 ignore_target_hosts, the router declines.
3407 7. Added expansion conditions match_domain, match_address, and match_local_
3408 part (NOT match_host).
3410 8. The placeholder for the Received: header didn't have a length field set.
3412 9. Added code to Exim itself and to exim_lock to test for a specific race
3413 condition that could lead to file corruption when using MBX delivery. The
3414 issue is with the lockfile that is created in /tmp. If this file is removed
3415 after a process has opened it but before that process has acquired a lock,
3416 there is the potential for a second process to recreate the file and also
3417 acquire a lock. This could lead to two Exim processes writing to the file
3418 at the same time. The added code performs the same test as UW imapd; it
3419 checks after acquiring the lock that its file descriptor still refers to
3420 the same named file.
3422 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3423 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3426 11. Added the smtp_active_hostname option. If used, this will typically be made
3427 to depend on the incoming interface address. Because $interface_address is
3428 not set up until the daemon has forked a reception process, error responses
3429 that can happen earlier (such as "too many connections") no longer contain
3432 12. If an expansion in a condition on a "warn" statement fails because a lookup
3433 defers, the "warn" statement is abandoned, and the next ACL statement is
3434 processed. Previously this caused the whole ACL to be aborted.
3436 13. Added the iplsearch lookup type.
3438 14. Added ident_timeout as a log selector.
3440 15. Added tls_certificate_verified as a log selector.
3442 16. Added a global option tls_require_ciphers (compare the smtp transport
3443 option of the same name). This controls incoming TLS connections.
3445 17. I finally figured out how to make tls_require_ciphers do a similar thing
3446 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3447 before starting the TLS session.
3449 18. Tabs are now shown as \t in -bP output.
3451 19. If the log selector return_path_on_delivery was set, Exim crashed when
3452 bouncing a message because it had too many Received: header lines.
3454 20. If two routers both had headers_remove settings, and the first one included
3455 a superfluous trailing colon, the final name in the first list and the
3456 first name in the second list were incorrectly joined into one item (with a
3457 colon in the middle).
3463 1. Added -C and -D options to the exinext utility, mainly to make it easier
3464 to include in the automated testing, but these could be helpful when
3465 multiple configurations are in use.
3467 2. The exinext utility was not formatting the output nicely when there was
3468 an alternate port involved in the retry record key, nor when there was a
3469 message id as well (for retries that were specific to a specific message
3470 and a specific host). It was also confused by IPv6 addresses, because of
3471 the additional colons they contain. I have fixed the IPv4 problem, and
3472 patched it up to do a reasonable job for IPv6.
3474 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3475 delivery, the log line now contains "pipelined" if PIPELINING was used.
3477 4. An SMTP transport process used to panic and die if the bind() call to set
3478 an explicit outgoing interface failed. This has been changed; it is now
3479 treated in the same way as a connect() failure.
3481 5. A reference to $sender_host_name in the part of a conditional expansion
3482 that was being skipped was still causing a DNS lookup. This no longer
3485 6. The def: expansion condition was not recognizing references to header lines
3486 that used bh_ and bheader_.
3488 7. Added the _cache feature to named lists.
3490 8. The code for checking quota_filecount in the appendfile transport was
3491 allowing one more file than it should have been.
3493 9. For compatibility with Sendmail, the command line option
3501 and sets the incoming protocol and host name (for trusted callers). The
3502 host name and its colon can be omitted when only the protocol is to be set.
3503 Note the Exim already has two private options, -pd and -ps, that refer to
3504 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3505 "s", but I don't think that's a major issue.
3507 10. A number of refactoring changes to the code, none of which should affect
3510 (a) The number of logging options was getting close to filling up the
3511 32-bit word that was used as a bit map. I have split them into two classes:
3512 those that are passed in the argument to log_write(), and those that are
3513 only ever tested independently outside of that function. These are now in
3514 separate 32-bit words, so there is plenty of room for expansion again.
3515 There is no change in the user interface or the logging behaviour.
3517 (b) When building, for example, log lines, the code previously used a
3518 macro that called string_cat() twice, in order to add two strings. This is
3519 not really sufficiently general. Furthermore, there was one instance where
3520 it was actually wrong because one of the argument was used twice, and in
3521 one call a function was used. (As it happened, calling the function twice
3522 did not affect the overall behaviour.) The macro has been replaced by a
3523 function that can join an arbitrary number of extra strings onto a growing
3526 (c) The code for expansion conditions now uses a table and a binary chop
3527 instead of a serial search (which was left over from when there were very
3528 few conditions). Also, it now recognizes conditions like "pam" even when
3529 the relevant support is not compiled in: a suitably worded error message is
3530 given if an attempt is made to use such a condition.
3532 11. Added ${time_interval:xxxxx}.
3534 12. A bug was causing one of the ddress fields not to be passed back correctly
3535 from remote delivery subprocesses. The field in question was not being
3536 subsequently used, so this caused to problems in practice.
3538 13. Added new log selectors queue_time and deliver_time.
3540 14. Might have fixed a bug in maildirsizefile handling that threw up
3541 "unexpected character" debug warnings, and recalculated the data
3542 unnecessarily. In any case, I expanded the warning message to give more
3545 15. Added the message "Restricted characters in address" to the statements in
3546 the default ACL that block characters like @ and % in local parts.
3548 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3549 Three changes have been made:
3551 (a) There was a serious bug; a negative response to MAIL caused the whole
3552 recipient domain to be cached as invalid, thereby blocking all messages
3553 to all local parts at the same domain, from all senders. This bug has
3554 been fixed. The domain is no longer cached after a negative response to
3555 MAIL if the sender used is not empty.
3557 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3560 (c) A new callout option, "use_sender" has been added for people who want
3561 the modified behaviour.
3567 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3570 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3571 indeed breaks things for older releases.
3573 3. Added additional logging to the case where there is a problem reading data
3574 from a filter that is running in a subprocess using a pipe, in order to
3575 try to track down a specific problem.
3577 4. Testing facility fudge: when running in the test harness and attempting
3578 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3579 getting "No route to host". Convert this to a timeout.
3581 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3584 6. Some OS don't have socklen_t but use size_t instead. This affects the
3585 fifth argument of getsockopt() amongst other things. This is now
3586 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3587 can be set for individual OS. I have set it for SunOS5, OSF1, and
3588 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3589 some earlier ones do not.
3591 7. Change 4.30/15 was not doing the test caselessly.
3593 8. The standard form for an IPv6 address literal was being rejected by address
3594 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3595 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3596 this, as well as the form without the "IPv6" on the front (but only when
3597 address literals are enabled, of course).
3599 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3601 10. Exim crashed if a message with an empty sender address specified by -f
3602 encountered a router with an errors_to setting. This could be provoked only
3603 by a command such as
3607 where an empty string was supplied; "<>" did not hit this bug.
3609 11. Installed PCRE release 4.5.
3611 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3612 remained set. It is now erased.
3614 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3615 times from message ids (which are base 36 rather than the normal 62).
3617 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3618 were being counted as actual protocol errors, and logged if the log
3619 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3620 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3621 and DATA following a set of rejected RCPTs do not count as protocol errors.
3622 In other words, Exim assumes they were pipelined, though this may not
3623 actually be the case. Of course, in all cases the client gets an
3624 appropriate error code.
3626 15. If a lookup fails in an ACL condition, a message about the failure may
3627 be available; it is used if testing the ACL cannot continue, because most
3628 such messages specify what the cause of the deferral is. However, some
3629 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3630 that caused an old message to be retained and used if a later statement
3631 caused a defer, replacing the real cause of the deferral.
3633 16. If an IP address had so many PTR records that the DNS lookup buffer
3634 was not large enough to hold them, Exim could crash while trying to process
3635 the truncated data. It now detects and logs this case.
3637 17. Further to 4.21/58, another change has been made: if (and only if) the
3638 first line of a message (the first header line) ends with CRLF, a bare LF
3639 in a subsequent header line has a space inserted after it, so as not to
3640 terminate the header.
3642 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3643 unnecessarily, used atoi() instead of strtol(), and didn't check the
3644 termination when getting file sizes from file names by regex.
3646 19. Completely re-implemented the support for maildirsize files, in the light
3647 of a number of problems with the previous contributed implementation
3648 (4.30/29). In particular:
3650 . If the quota is zero, the maildirsize file is maintained, but no quota is
3653 . If the maildir directory does not exist, it is created before any attempt
3654 to write a maildirsize file.
3656 . The quota value in the file is just a cache; if the quota is changed in
3657 the transport, the new value overrides.
3659 . A regular expression is available for excluding directories from the
3662 20. The autoreply transport checks the characters in options that define the
3663 message's headers; it allows continued headers, but it was checking with
3664 isspace() after an embedded newline instead of explicitly looking for a
3667 21. If all the "regular" hosts to which an address was routed had passed their
3668 expiry times, and had not reached their retry times, the address was
3669 bounced, even if fallback hosts were defined. Now Exim should go on to try
3672 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3673 equivalent code in the SMTP transport. Some hosts send humungous responses
3674 to HELO/EHLO, more than 1024 it seems.
3676 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3677 gives compiler warnings in some environments. I've now done it "properly",
3680 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3681 (because of problems with the built-in one) was declared to return uschar *
3682 instead of char *, causing compiler failure.
3684 25. Fixed a file descriptor leak when processing alias/forward files.
3686 26. Fixed a minor format string issue in dbfn.c.
3688 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3690 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3691 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3694 29. When a sender address is verified, it is cached, to save repeating the test
3695 when there is more than one recipient in a message. However, when the
3696 verification involves a callout, it is possible for different callout
3697 options to be set for different recipients. It is too complicated to keep
3698 track of this in the cache, so now Exim always runs a verification when a
3699 callout is required, relying on the callout cache for the optimization.
3700 The overhead is duplication of the address routing, but this should not be
3703 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3704 to be verified with callout=postmaster, and the main callout worked but the
3705 postmaster check failed, the verification correctly failed. However, if a
3706 subsequent RCPT command asked for sender verification *without* the
3707 postmaster check, incorrect caching caused this verification also to fail,
3710 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3711 it was not caching the DNS options (qualify_single, search_parents) that
3712 were used when the lookup failed. A subsequent lookup with different
3713 options therefore always gave the same answer, though there were cases
3714 where it should not have. (Example: a "domains = !$mx_any" option on a
3715 dnslookup router: the "domains" option is always processed without any
3716 widening, but the router might have qualify_single set.) Now Exim uses the
3717 cached value only when the same options are set.
3719 32. Added John Jetmore's "exipick" utility to the distribution.
3721 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3722 than a timeout (e.g. a certificate is required, and is not provided), an
3723 Exim server now closes the connection immediately. Previously it waited for
3724 the client to close - but if the client is SSL, it seems that they each
3725 wait for each other, leading to a delay before one of them times out.
3727 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3728 maintained 0.8.x compatibility because I don't think many are using it, and
3729 it is clearly obsolete.
3731 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3734 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3735 client certificate was expired. A simple patch fixes this, though I don't
3736 understand the full logic of why the verify callback is called multiple
3739 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3740 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3741 which causes problems with some clients (such as the Certicom SSL Plus
3742 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3743 disables the coutermeasure allowing Eudora to connect."
3745 38. Exim was not checking that a write() to a log file succeeded. This could
3746 lead to Bad Things if a log got too big, in particular if it hit a file
3747 size limit. Exim now panics and dies if it cannot write to a log file, just
3748 as it does if it cannot open a log file.
3750 39. Modified OS/Makefile-Linux so that it now contains
3752 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3754 The two -D definitions ensure that Exim is compiled with large file
3755 support, which makes it possible to handle log files that are bigger than
3758 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3759 instance) a domain was checked against a named list that involved a lookup,
3760 causing $domain_data to be set, then another domain was checked against the
3761 same list, then the first domain was re-checked, the value of $domain_data
3762 after the final check could be wrong. In particular, if the second check
3763 failed, it could be set empty. This bug probably also applied to
3766 41. The strip_trailing_dot option was not being applied to the address given
3767 with the -f command-line option.
3769 42. The code for reading a message's header from the spool was incrementing
3770 $received_count, but never initializing it. This meant that the value was
3771 incorrect (doubled) while delivering a message in the same process in which
3772 it was received. In the most common configuration of Exim, this never
3773 happens - a fresh exec is done - but it can happen when
3774 deliver_drop_privilege is set.
3776 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3777 it now includes up to 150 characters of the unexpected data in the log
3780 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3781 and building data strings. The size of both of these buffers was 10 000
3782 bytes - far larger than anybody would *ever* want, thought I. Needless to
3783 say, somebody hit the limit. I have increased the maximum line length to
3784 20 000 and the maximum data length of concatenated lines to 100 000. I have
3785 also fixed two bugs, because there was no checking on these buffers. Tsk,
3786 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3787 buffer is too small.
3789 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3790 lsearch lookups. Now it does.
3792 46. When parsing a route_list item in a manualroute router, a fixed-length
3793 buffer was used for the list of hosts. I made this 1024 bytes long,
3794 thinking that nobody would ever have a list of hosts that long. Wrong.
3795 Somebody had a whole pile of complicated expansion conditions, and the
3796 string was silently truncated, leading to an expansion error. It turns out
3797 that it is easier to change to an unlimited length (owing to other changes
3798 that have happened since this code was originally written) than to build
3799 structure for giving a limitation error. The length of the item that
3800 expands into the list of hosts is now unlimited.
3802 47. The lsearch lookup could not handle data where the length of text line was
3803 more than 4095 characters. Such lines were truncated, leading to shortened
3804 data being returned. It should now handle lines of any length.
3806 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3807 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3810 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3811 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3812 no longer happens in comment lines. A list of replacements is now placed
3813 at the head of all of the source files, except those whose only change is
3814 to replace PERL_COMMAND in the very first #! line.
3816 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3817 messages on the queue, with a bottom-up merge sort, using code contributed
3818 by Michael Haardt. This should make operations like -bp somewhat faster on
3819 large queues. It won't affect queue runners, except when queue_run_in_order
3822 51. Installed eximstats 1.31 in the distribution.
3824 52. Added support for SRV lookups to the dnslookup router.
3826 53. If an ACL referred to $message_body or $message_body_end, the value was not
3827 reset for any messages that followed in the same SMTP session.
3829 54. The store-handling optimization for building very long strings was not
3830 differentiating between the different store pools. I don't think this
3831 actually made any difference in practice, but I've tidied it.
3833 55. While running the routers to verify a sender address, $sender_address
3834 was still set to the sender address. This is wrong, because when routing to
3835 send a bounce to the sender, it would be empty. Therefore, I have changed
3836 it so that, while verifying a sender address, $sender_address is set to <>.
3837 (There is no change to what happens when verifying a recipient address.)
3839 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3840 target A or AAAA records (if not already returned) without resetting the
3841 qualify_single or search_parents options of the DNS resolver. These are
3842 inappropriate in this case because the targets of MX and SRV records must
3843 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3844 target that, when qualified, matched something in the local domain. These
3845 two options are now turned off when doing these lookups.
3847 57. It seems that at least some releases of Reiserfs (which does not have the
3848 concept of a fixed number of inodes) returns zero and not -1 for the
3849 number of available inodes. This interacted badly with check_spool_inodes,
3850 which assumed that -1 was the "no such thing" setting. What I have done is
3851 to check that the total number of inodes is greater than zero before doing
3852 the test of how many are available.
3854 58. When a "warn" ACL statement has a log_message modifier, the message is
3855 remembered, and not repeated. This is to avoid a lot of repetition when a
3856 message has many recipients that cause the same warning to be written.
3857 Howewer, Exim was preserving the list of already written lines for an
3858 entire SMTP session, which doesn't seem right. The memory is now reset if a
3859 new message is started.
3861 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3862 debugging output unless log_rewrite was also set.
3864 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3865 of (int)(handle) when we know that handle contains (void *)(-1).
3867 61. The Exim daemon panic-logs an error return when it closes the incoming
3868 connection. However "connection reset by peer" seems to be common, and
3869 isn't really an error worthy of noting specially, so that particular error
3872 62. When Exim is trying to find all the local interfaces, it used to panic and
3873 die if the ioctl to get the interface flags failed. However, it seems that
3874 on at least one OS (Solaris 9) it is possible to have an interface that is
3875 included in the list of interfaces, but for which you get a failure error
3876 for this call. This happens when the interface is not "plumbed" into a
3877 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3878 failure of the "get flags" call assumes that the interface is down.
3880 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3881 makes life easier for people who are doing arithmetic on fields extracted
3882 from dates, where you often get leading zeros that should not be
3883 interpreted as octal.
3885 64. Added qualify_domain to the redirect router, to override the global
3888 65. If a pathologically long header line contained very many addresses (the
3889 report of this problem mentioned 10 000) and each of them was rewritten,
3890 Exim could use up a very large amount of memory. (It kept on making new
3891 copies of the header line as it rewrote, and never released the old ones.)
3892 At the expense of a bit more processing, the header rewriting function has
3893 been changed so that it no longer eats memory in this way.
3895 66. The generation of the Received: header has been moved from the time that a
3896 message starts to be received, to the time that it finishes. The timestamp
3897 in the Received: header should now be very close to that of the <= log
3898 line. There are two side-effects of this change:
3900 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3901 logged header lines no longer include the local Received: line, because
3902 it has not yet been created. The same applies to a copy of the message
3903 that is returned to a non-SMTP sender when a message is rejected.
3905 (b) When a filter file is tested using -bf, no additional Received: header
3906 is added to the test message. After some thought, I decided that this
3909 This change does not affect the value of $received_for. It is still set
3910 after address rewriting, but before local_scan() is called.
3912 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3914 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3915 gave an unhelpful panic error message, and a defer error. I have managed to
3916 change this behaviour so that it now rejects any supplied certificate,
3917 which seems right, as the list of acceptable certificates is empty.
3919 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3920 gave an unhelpful defer error. I have not managed to make this reject any
3921 supplied certificates, but the error message it gives is "no certificate
3922 supplied", which is not helpful.
3924 70. exigrep's output now also includes lines that are not associated with any
3925 message, but which match the given pattern. Implemented by a patch from
3926 Martin Sluka, which also tidied up the Perl a bit.
3928 71. Recipient callout verification, like sender verification, was using <> in
3929 the MAIL FROM command. This isn't really the right thing, since the actual
3930 sender may affect whether the remote host accepts the recipient or not. I
3931 have changed it to use the actual sender in the callout; this means that
3932 the cache record is now keyed on a recipient/sender pair, not just the
3933 recipient address. There doesn't seem to be a real danger of callout loops,
3934 since a callout by the remote host to check the sender would use <>.
3935 [SEE ABOVE: changed after hitting problems.]
3937 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3938 temporary errors. However, in the case of such a code being given after
3939 the end of a data transmission (i.e. after ".") Exim was failing to write
3940 a retry record for the message. (Yes, there was some broken host that was
3941 actually sending 8xx at this point.)
3943 73. An unknown lookup type in a host list could cause Exim to panic-die when
3944 the list was checked. (An example that provoked this was putting <; in the
3945 middle of a list instead of at the start.) If this happened during a DATA
3946 ACL check, a -D file could be left lying around. This kind of configuration
3947 error no longer causes Exim to die; instead it causes a defer errror. The
3948 incident is still logged to the main and panic logs.
3950 74. Buglet left over from Exim 3 conversion. The message "too many messages
3951 in one connection" was written to the rejectlog but not the mainlog, except
3952 when address rewriting (yes!) was being logged.
3954 75. Added write_rejectlog option.
3956 76. When a system filter was run not as root (that is, when system_filter_user
3957 was set), the values of the $n variables were not being returned to the
3958 main process; thus, they were not subsequently available in the $sn
3961 77. Added +return_path_on_delivery log selector.
3963 78. A connection timeout was being treated differently from recipients deferred
3964 when testing hosts_max_try with a message that was older than the host's
3965 retry timeout. (The host should not be counted, thus allowing all hosts to
3966 be tried at least once before bouncing.) This may have been the cause of an
3967 occasionally reported bug whereby a message would remain on the queue
3968 longer than the retry timeout, but would be bounced if a delivery was
3969 forced. I say "may" because I never totally pinned down the problem;
3970 setting up timeout/retry tests is difficult. See also the next item.
3972 79. The ultimate address timeout was not being applied to errors that involved
3973 a combination of host plus message (for example, a timeout on a MAIL
3974 command). When an address resolved to a number of possible hosts, and they
3975 were not all tried for each delivery (e.g. because of hosts_max_try), a
3976 message could remain on the queue longer than the retry timeout.
3978 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3981 81. Fixed an obscure SMTP outgoing bug which required at least the following
3982 conditions: (a) there was another message waiting for the same server;
3983 (b) the server returned 5xx to all RCPT commands in the first message so
3984 that the message was not completed; (c) the server dropped the connection
3985 or gave a negative response to the RSET that Exim sends to abort the
3986 transaction. The observed case was a dropped connection after DATA that had
3987 been sent in pipelining mode. That is, the server had advertised PIPELINING
3988 but was not implementing it correctly. The effect of the bug was incorrect
3989 behaviour, such as trying another host, and this could lead to a crash.
3995 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3996 and daemon.c were passed as pointers to ints; they should have been
3997 pointers to socklen_t variables (which are typically unsigned ints).
3999 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4002 3. Fixed a really odd bug that affected only the testing scheme; patching a
4003 certain fixed string in the binary changed the value of another string that
4004 happened to be identical to the end of the original first string.
4006 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4007 name", it returns that address as the IP address. On some operating
4008 systems (e.g. Solaris), it also passes back the IP address string as the
4009 "host name". However, on others (e.g. Linux), it passes back an empty
4010 string. Exim wasn't checking for this, and was changing the host name to an
4011 empty string, assuming it had been canonicized.
4013 5. Although rare, it is permitted to have more than one PTR record for a given
4014 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4015 all the names associated with an address, because they do in Solaris.
4016 However, it seems that they do not in Linux for data that comes from the
4017 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4018 I found this out when I moved to a new Linux workstation and tried to run
4019 the Exim test suite.
4021 To get round this problem I have changed the code so that it now does its
4022 own call to the DNS to look up PTR records when searching for a host name.
4023 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4024 addresses that are only in /etc/hosts are still found.
4026 This behaviour is, however, controlled by an option called host_lookup_
4027 order, which defaults to "bydns:byaddr". If people want to use the other
4028 order, or indeed, just use one or the other means of lookup, they can
4029 specify it in this variable.
4031 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4032 some operating systems, this comes back from gethostbyaddr() as an empty
4033 string, and this is what Exim used to test for. However, it seems that in
4034 other systems, "." is yielded. Exim now tests for this case too.
4036 7. The values of check_spool_space and check_log_space are now held internally
4037 as a number of kilobytes instead of an absolute number of bytes. If a
4038 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4039 kilobyte. This means that much larger values can be stored.
4041 8. Exim monitor: an attempt to get the action menu when not actually pointing
4042 at a message produces an empty menu entitled "No message selected". This
4043 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4044 no entries in it ("Shell widget menu has zero width and/or height"). So I
4045 have added a single, blank menu entry in this case.
4047 9. Added ${quote_local_part.
4049 10. MIME decoding is now applied to the contents of Subject: header lines when
4052 11. Now that a reference to $sender_host_address automatically causes a reverse
4053 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4054 host lookup before query-style lookups in lists that might use this
4055 variable. This has therefore been abolished, and the "net-" prefix is no
4056 longer necessary for query-style lookups.
4058 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4059 have been a typo for LFLAGS, so it has been changed.
4061 13. The install script calls Exim with "-C /dev/null" in order to find the
4062 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4063 to be output. Howeve, since Exim outputs its version number before the
4064 error, it didn't break the script. It just looked ugly. I fixed this by
4065 always allowing "-C /dev/null" if the caller is root.
4067 14. Ignore overlarge ACL variable number when reading spool file - insurance
4068 against a later release with more variables having written the file.
4070 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4071 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4072 this, as well as the form without the "IPv6" on the front.
4074 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4075 OS/Makefile-Darwin file.
4077 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4078 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4080 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4081 variable. However, some OS use ints for inodes. I've added cast to long int
4082 to get rid of the compiler warning.
4084 19. I had forgotten to lock out "/../" in configuration file names when
4085 ALT_CONFIG_PREFIX was set.
4087 20. Routers used for verification do not need to specify transports. However,
4088 if such a router generated a host list, and callout was configured, Exim
4089 crashed, because it could not find a port number from the (non-existent)
4090 transport. It now assumes port 25 in this circumstance.
4092 21. Added the -t option to exigrep.
4094 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4095 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4096 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4097 them is set, LOOKUP_LSEARCH is forced.
4099 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4101 24. Added sender and host information to the "rejected by local_scan()" log
4102 line; previously there was no indication of these.
4104 25. Added .include_if_exists.
4106 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4107 Linux. It turns out that not all file systems support this. Apparently some
4108 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4109 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4110 not supported on the file descriptor, is now ignored when Exim is trying to
4111 sync a directory. This applies only to Linux.
4113 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4115 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4116 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4119 29. Added support for maildirsize files from supplied patch (modified a bit).
4121 30. The use of :fail: followed by an empty string could lead Exim to respond to
4122 sender verification failures with (e.g.):
4124 550 Verification failed for <xxx>
4125 550 Sender verify failed
4127 where the first response line was missing the '-' that indicates it is not
4128 the final line of the response.
4130 31. The loop for finding the name of the user that called Exim had a hardwired
4131 limit of 10; it now uses the value of finduser_retries, which is used for
4132 all other user lookups.
4134 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4137 33. Exim was neglecting to zero errno before one call of strtol() when
4138 expanding a string and expecting an integer value. On some systems this
4139 resulted in spurious "integer overflow" errors. Also, it was casting the
4140 result into an int without checking.
4142 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4143 did not work. The code looks as if it has *never* worked, though it appears
4144 to have been documented since at least releast 1.62. I have made it work.
4146 35. The "timeout_DNS" error in retry rules, also documented since at least
4147 1.62, also never worked. As it isn't clear exactly what this means, and
4148 clearly it isn't a major issue, I have abolished the feature by treating it
4149 as "timeout", and writing a warning to the main and panic logs.
4151 36. The display of retry rules for -brt wasn't always showing the error code
4154 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4155 timeout_connect_A, timeout_connect_MX.
4157 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4158 to the empty sender.
4160 39. The daemon was not analysing the content of -oX till after it had closed
4161 stderr and disconnected from the controlling terminal. This meant that any
4162 syntax errors were only noted on the panic log, and the return code from
4163 the command was 0. By re-arranging the code a little, I've made the
4164 decoding happen first, so such errors now appear on stderr, and the return
4165 code is 1. However, the actual setting up of the sockets still happens in
4166 the disconnected process, so errors there are still only recorded on the
4169 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4170 connections (as happens on some IP stacks) was logged at start up time as
4171 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4172 it from "IPv6 and IPv4", which means that two separate sockets are being
4175 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4176 says whether AF_INET or AF_INET6 was passed as an argument.
4178 42. Exiwhat output was messed up when time zones were included in log
4181 43. Exiwhat now gives more information about the daemon's listening ports,
4182 and whether -tls-on-connect was used.
4184 44. The "port" option of the smtp transport is now expanded.
4186 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4187 silently ignored. Now an error message is written to the main and panic
4190 46. There's a new ACL modifier called "logwrite" which writes to a log file
4191 as soon as it is encountered.
4193 47. Added $local_user_uid and $local_user_gid at routing time.
4195 48. Exim crashed when trying to verify a sender address that was being
4198 49. Exim was recognizing only a space character after ".include". It now also
4199 recognizes a tab character.
4201 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4202 extracting the relevant information from the specification. The man page no
4203 longer contains scrambled data for the -d option, and I've added a section
4204 at the front about calling Exim under different names.
4206 51. Added "extra_headers" argument to the "mail" command in filter files.
4208 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4211 53. Installed eximstats 1.29.
4213 54. Added transport_filter_timeout as a generic transport option.
4215 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4216 Cc: header lines. This was required by RFC 822, but it not required by RFC
4219 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4220 incoming messages that did not have them. Now it does so only if the
4221 message originates locally, that is, if there is no associated remote host
4222 address. When Resent- header lines are present, this applies to the Resent-
4223 lines rather than the non-Resent- lines.
4225 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4226 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4228 58. Messages for configuration errors now include the name of the main
4229 configuration file - useful now that there may be more than one file in a
4230 list (.included file names were always shown).
4232 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4233 for those rare installations that do not start the daemon as root or run it
4234 setuid root. I've cut out the call to initgroups() if the daemon is not
4237 60. The Exim user and group can now be bound into the binary as text strings
4238 that are looked up at the start of Exim's processing.
4240 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4242 62. Added $mailstore_basename variable.
4244 63. Installed patch to sieve.c from Michael Haardt.
4246 64. When Exim failed to open the panic log after failing to open the main log,
4247 the original message it was trying to log was written to stderr and debug
4248 output, but if they were not available (the usual case in production), it
4249 was lost. Now it is written to syslog before the two lines that record the
4250 failures to open the logs.
4252 65. Users' Exim filters run in subprocesses under the user's uid. It is
4253 possible for a "deliver" command or an alias in a "personal" command to
4254 provoke an address rewrite. If logging of address rewriting is configured,
4255 this fails because the process is not running as root or exim. There may be
4256 a better way of dealing with this, but for the moment (because 4.30 needs
4257 to be released), I have disabled address rewrite logging when running a
4258 filter in a non-root, non-exim process.
4264 1. The buildconfig auxiliary program wasn't quoting the value set for
4265 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4266 not defined. This bug was masked in 4.22 by the effect that was fixed in
4269 2. Some messages that were rejected after a message id was allocated were
4270 shown as "incomplete" by exigrep. It no longer does this for messages that
4271 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4273 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4274 have allow_domain_literals set, the ID did not get logged in the <= line.
4275 Domain literals are now always recognized in Message-ID: header lines.
4277 4. The first argument for a ${extract expansion item is the key name or field
4278 number. Leading and trailing spaces in this item were not being ignored,
4279 causing some misleading effects.
4281 5. When deliver_drop_privilege was set, single queue runner processes started
4282 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4283 same command in the process it spins off) were not dropping privilege.
4285 6. When the daemon running as "exim" started a queue runner, it always
4286 re-executed Exim in the spun-off process. This is a waste of effort when
4287 deliver_drop_privilege is set. The new process now just calls the
4288 queue-runner function directly.
4294 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4297 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4298 ignored. Though the use of -oP was forcing the writing of a pid file, it
4299 was always written to the default place.
4301 3. If the message "no IP address found for host xxxx" is generated during
4302 incoming verification, it is now followed by identification of the incoming
4303 connection (so you can more easily find what provoked it).
4305 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4307 5. Added some features to "harden" Exim a bit more against certain attacks:
4309 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4310 be put in Local/Makefile. This is like the never_users runtime option,
4311 but it cannot be overridden. The default setting is "root".
4313 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4314 prefix string with which any file named in a -C command line option
4317 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4318 is retained for -C and -D only if the caller of Exim is root. Without
4319 it, the exim user may also use -C and -D and retain privilege.
4321 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4322 command line option is disabled.
4324 6. Macro names set by the -D option must start with an upper case letter, just
4325 like macro names defined in the configuration file.
4327 7. Added "dereference=" facility to LDAP.
4329 8. Two instances of the typo "uknown" in the source files are fixed.
4331 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4332 the Configure-Makefile script screwed up while processing it.
4334 10. Incorporated PCRE 4.4.
4336 11. The SMTP synchronization check was not operating right at the start of an
4337 SMTP session. For example, it could not catch a HELO sent before the client
4338 waited for the greeting. There is now a check for outstanding input at the
4339 point when the greeting is written. Because of the duplex, asynchronous
4340 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4341 way, but not yet received, when the check is performed.
4343 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4344 on TCP/IP sockets, because this apparently causes some broken clients to
4347 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4348 unchanged) from the Cygwin maintainer.
4350 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4351 was supported instead of testing for mbx. Effectively a typo.
4353 15. The spa authenticator server code was not checking that the input it
4354 received was valid base64.
4356 16. The debug output line for the "set" modifier in ACLs was not showing the
4357 name of the variable that was being set.
4359 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4361 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4362 lookup on its own. Something else was needed to trigger the lookup. For
4363 example, a match in host_lookup or the need for a host name in a host list.
4364 Now, if $sender_host_name is referenced and the host name has not yet been
4365 looked up, a lookup is performed. If the lookup fails, the variable remains
4366 empty, and $host_lookup_failed is set to "1".
4368 19. Added "eqi" as a case-independent comparison operator.
4370 20. The saslauthd authentication condition could segfault if neither service
4371 nor realm was specified.
4373 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4374 error message that was logged was misleading, and incoming SMTP
4375 connections were dropped. The message is now more accurate, and temporary
4376 errors are given to SMTP connections.
4378 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4379 (except RSET) with a 421 error, until QUIT is received. However, it was
4380 failing to send a response to QUIT.
4382 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4383 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4384 ACL is now run after the helo_try_verify_hosts code.
4386 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4387 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4388 case-independent, so other case variants are also recognized). Apparently
4389 some systems use these upper case variants.
4391 25. If more than two messages were waiting for the same host, and a transport
4392 filter was specified for the transport, Exim sent two messages over the
4393 same TCP/IP connection, and then failed with "socket operation on non-
4394 socket" when it tried to send the third.
4396 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4398 27. The extern definition of crypt16() in expand.c was not being excluded when
4399 the OS had its own crypt16() function.
4401 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4402 as a preferred synonym for return_size_limit, both as an option and as an
4405 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4407 30. Changed the default configuration ACL to relax the local part checking rule
4408 for addresses that are not in any local domains. For these addresses,
4409 slashes and pipe symbols are allowed within local parts, but the sequence
4410 /../ is explicitly forbidden.
4412 31. SPA server authentication was not clearing the challenge buffer before
4415 32. log_message in a "warn" ACL statement was writing to the reject log as
4416 well as to the main log, which contradicts the documentation and doesn't
4417 seem right (because no rejection is happening). So I have stopped it.
4419 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4420 However, I am unable to do any testing of this.
4422 34. Fixed an infelicity in the appendfile transport. When checking directories
4423 for a mailbox, to see if any needed to be created, it was accidentally
4424 using path names with one or more superfluous leading slashes; tracing
4425 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4427 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4428 contents are added to the log line that is written for every discarded
4429 recipient. (Previously a log_message setting was ignored.)
4431 36. The ${quote: operator now quotes the string if it is empty.
4433 37. The install script runs exim in order to find its version number. If for
4434 some reason other than non-existence or emptiness, which it checks, it
4435 could not run './exim', it was installing it with an empty version number,
4436 i.e. as "exim-". This error state is now caught, and the installation is
4439 38. An argument was missing from the function that creates an error message
4440 when Exim fails to connect to the socket for saslauthd authentication.
4441 This could cause Exim to crash, or give a corrupted message.
4443 39. Added isip, isip4, and isip6 to ${if conditions.
4445 40. The ACL variables $acl_xx are now saved with the message, and can be
4446 accessed later in routers, transports, and filters.
4448 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4449 strings in the file are not string-expanded.
4451 42. If a MAIL command specified a SIZE value that was too large to fit into an
4452 int variable, the check against message_size_limit failed. Such values are
4453 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4454 day this will have to be increased, but I don't think I want to be around
4455 when emails are that large.
4462 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4463 iconv() is not standard in FreeBSD.
4465 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4466 IPv6 enabled. The observed symptom was a segmentation fault on return from
4467 the function os_common_find_running_interfaces() in src/os.c.
4469 3. In the check_special_case() function in daemon.c I had used "errno" as an
4470 argument name, which causes warnings on some systems. This was basically a
4471 typo, since it was named "eno" in the comments!
4473 4. The code that waits for the clock to tick (at a resolution of some fraction
4474 of a second) so as to ensure message-id uniqueness was always waiting for
4475 at least one whole tick, when it could have waited for less. [This is
4476 almost certainly not relevant at current processor speeds, where it is
4477 unlikely to ever wait at all. But we try to future-proof.]
4479 5. The function that sleeps for a time interval that includes fractions of a
4480 second contained a race. It did not block SIGALRM between setting the
4481 timer, and suspending (a couple of lines later). If the interval was short
4482 and the sigsuspend() was delayed until after it had expired, the suspension
4483 never ended. On busy systems this could lead to processes getting stuck for
4486 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4487 process, before it forks any delivery processes. The open lookup caching
4488 mechanism meant that the open file or database connection was passed into
4489 the delivery process. The problem was that delivery processes always tidy
4490 up cached lookup data. This could cause a problem for the next delivery
4491 process started by the queue runner, because the external queue runner
4492 process does not know about the closure. So the next delivery process
4493 still has data in the lookup cache. In the case of a file lookup, there was
4494 no problem because closing a file descriptor in a subprocess doesn't affect
4495 the parent. However, if the lookup was caching a connection to a database,
4496 the connection was closed, and the second delivery process was likely to
4497 see errors such as "PGSQL: query failed: server closed the connection
4498 unexpectedly". The problem has been fixed by closing all cached lookups
4499 in a queue runner before running a delivery process.
4501 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4502 seem to have the "const" qualifier which it has on other OS. I've
4505 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4506 *of the same type* (client or server) with the same public name that an
4507 error should be diagnosed.
4509 9. When Exim looked up a host name for an IP address, but failed to find the
4510 original IP address when looking up the host name (a safety check), it
4511 output the message "<ip address> does not match any IP for NULL", which was
4512 confusing, to say the least. The bug was that the host name should have
4513 appeared instead of "NULL".
4515 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4516 user that is built into the binary, and the -C or -D options is used, root
4517 privilege is dropped before the configuration file is read. In addition,
4518 logging is switched to stderr instead of the normal log files. If the
4519 configuration then re-defines the Exim user, the unprivileged environment
4520 is probably not what is expected, so Exim logs a panic warning message (but
4523 However, if deliver_drop_privilege is set, the unprivileged state may well
4524 be exactly what is intended, so the warning has been cut out in that case,
4525 and Exim is allowed to try to write to its normal log files.
4531 1. smtp_return_error_details was not giving details for temporary sender
4532 or receiver verification errors.
4534 2. Diagnose a configuration error if two authenticators have the same public
4537 3. Exim used not to create the message log file for a message until the first
4538 delivery attempt. This could be confusing when incoming messages were held
4539 for policy or load reasons. The message log file is now created at the time
4540 the message is received, and an initial "Received" line is written to it.
4542 4. The automatically generated man page for command line options had a minor
4543 bug that caused no ill effects; however, a more serious problem was that
4544 the procedure for building the man page automatically didn't always
4545 operate. Consequently, release 4.20 contains an out-of-date version. This
4546 shouldn't happen again.
4548 5. When building Exim with embedded Perl support, the script that builds the
4549 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4550 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4552 6. The freeze_tell option was not being used for messages that were frozen on
4553 arrival, either by an ACL or by local_scan().
4555 7. Added the smtp_incomplete_transaction log selector.
4557 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4558 was accepting AUTH without a new EHLO.
4560 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4561 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4564 10. Logging of TCP/IP connections (when configured) now happens in the main
4565 daemon process instead of the child process, so that the TCP/IP connection
4566 count is more accurate (but it can never be perfect).
4568 11. The use of "drop" in a nested ACL was not being handled correctly in the
4569 outer ACL. Now, if condition failure induced by the nested "drop" causes
4570 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4571 or "require"), the connection is dropped.
4573 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4574 that yield "discard" can now be used with an "accept" or a "discard" verb,
4575 but an error is generated for any others (because I can't see a useful way
4576 to define what should happen).
4578 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4579 are now processed in the same way as lines in the Exim configuration file.
4580 In particular, continuation lines are supported.
4582 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4584 15. Added -ti meaning -t -i.
4586 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4587 domains, and warn by logging if others are found.
4589 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4590 returned from GIFCONF when Exim is trying to find the list of interfaces on
4591 a host. The code in os.c has been modified to copy each ifreq to an aligned
4592 structure in all cases.
4594 Also, in some cases, the returned ifreq's were being copied to a 'struct
4595 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4596 means the last couple of bytes of an IPv6 address could be chopped if the
4597 ifreq contained only a normal sockaddr (14 bytes storage).
4599 18. Named domain lists were not supported in the hosts_treat_as_local option.
4600 An entry such as +xxxx was not recognized, and was treated as a literal
4603 19. Ensure that header lines added by a DATA ACL are included in the reject log
4604 if the ACL subsequently rejects the message.
4606 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4607 MD5 (which is deprecated).
4609 21. When testing a filter file using -bf, Exim was writing a message when it
4610 took the sender from a "From " line in the message, but it was not doing so
4611 when it took $return_path from a Return-Path: header line. It now does.
4613 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4614 with a valid header line field name (a series of printing characters
4615 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4617 23. Changed "disc" in the source to "disk" to conform to the documentation and
4618 the book and for uniformity.
4620 24. Ignore Sendmail's -Ooption=value command line item.
4622 25. When execve() failed while trying to run a command in a pipe transport,
4623 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4624 could be confused with a return value of 69 from the command itself. This
4625 has been changed to 127, the value the shell returns if it is asked to run
4626 a non-existent command. The wording for the related log line suggests a
4627 non-existent command as the problem.
4629 26. If received_header_text expands to an empty string, do not add a Received:
4630 header line to the message. (Well, it adds a token one on the spool, but
4631 marks it "old" so that it doesn't get used or transmitted.)
4633 27. Installed eximstats 1.28 (addition of -nt option).
4635 28. There was no check for failure on the call to getsockname() in the daemon
4636 code. This can fail if there is a shortage of resources on the system, with
4637 ENOMEM, for example. A temporary error is now given on failure.
4639 29. Contrary to the C standard, it seems that in some environments, the
4640 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4641 program. Exim now does this explicitly; it affects the formatting of
4642 timestamps using strftime().
4644 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4645 complaints. I've now initialized all the variables, to avoid this.
4647 32. Header lines added by a system filter were not being "seen" during
4648 transport-time rewrites.
4650 33. The info_callback() function passed to OpenSSL is set up with type void
4651 (*)(SSL *, int, int), as described somewhere. However, when calling the
4652 function (actually a macro) that sets it up, the type void(*)() is
4653 expected. I've put in a cast to prevent warnings from picky compilers.
4655 34. If a DNS black list lookup found a CNAME record, but there were no A
4656 records associated with the domain it pointed at, Exim crashed.
4658 35. If a DNS black list lookup returned more than one A record, Exim ignored
4659 all but the first. It now scans all returned addresses if a particular IP
4660 value is being sought. In this situation, the contents of the
4661 $dnslist_value variable are a list of all the addresses, separated by a
4664 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4665 to accept a host name and all its aliases if the forward lookup for any of
4666 them yielded the IP address of the incoming connection. Now it accepts only
4667 those names whose forward lookup yields the correct IP address. Any other
4668 names are discarded. This closes a loophole whereby a rogue DNS
4669 administrator could create reverse DNS records to break through a
4670 wildcarded host restriction in an ACL.
4672 37. If a user filter or a system filter that ran in a subprocess used any of
4673 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4674 the wrong values were passed to the pipe command ($thisaddress had the
4675 value of $0, $0 had the value of $1, etc). This bug was introduced by
4676 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4678 38. Improved the line breaking for long SMTP error messages from ACLs.
4679 Previously, if there was no break point between 40 and 75 characters, Exim
4680 left the rest of the message alone. Two changes have been made: (a) I've
4681 reduced the minimum length to 35 characters; (b) if it can't find a break
4682 point between 35 and 75 characters, it looks ahead and uses the first one
4683 that it finds. This may give the occasional overlong line, but at least the
4684 remaining text gets split now.
4686 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4687 file descriptors might be low, and that setting 1000 would always raise it.
4688 It turns out that in some environments, the limit is already over 1000 and
4689 that lowering it causes trouble. So now Exim takes care not to decrease it.
4691 40. When delivering a message, the value of $return_path is set to $sender_
4692 address at the start of routing (routers may change the value). By an
4693 oversight, this default was not being set up when an address was tested by
4694 -bt or -bv, which affected the outcome if any router or filter referred to
4697 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4698 log only when "message" or "log_message" are set. However, if one of the
4699 conditions was an address verification, or a call to a nested ACL, the
4700 messages generated by the underlying test were being passed through. This
4701 no longer happens. The underlying message is available in $acl_verify_
4702 message for both "message" and "log_message" expansions, so it can be
4703 passed through if needed.
4705 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4706 new expansion $bh_ to give the encoded byte string without charset
4707 translation. Translation happens only if iconv() is available; HAVE_ICONV
4708 indicates this at build time. HEADERS_CHARSET gives the charset to
4709 translate to; headers_charset can change it in the configuration, and
4710 "headers charset" can change it in an individual filter file.
4712 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4713 that creates RFC 2047 encoded "words" labels them as that charset instead
4714 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4715 expansion operator; (ii) when Exim creates a From: line for a local
4716 message; (iii) when a header line is rewritten to include a "phrase" part.
4718 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4719 buggy, causing it to skip the first lines of messages whose message ID
4720 ended in 'D'. This would not have bitten before Exim release 4.14, because
4721 message IDs were unlikely to end in 'D' before then. The effect was to have
4722 incorrect size information for certain domains.
4724 45. #include "config.h" was missing at the start of the crypt16.c module. This
4725 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4728 46. If there was a timeout during a "random" callout check, Exim treated it as
4729 a failure of the random address, and carried on sending RSET and the real
4730 address. If the delay was just some slowness somewhere, the response to the
4731 original RCPT would be taken as a response to RSET and so on, causing
4732 mayhem of various kinds.
4734 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4735 when I implemented it. It didn't allow for the fact that some option values
4736 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4737 the right test for positive values.
4739 48. Domain names in DNS records are case-independent. Exim always looks them up
4740 in lower case. Some resolvers return domain names in exactly the case they
4741 appear in the zone file, that is, they may contain uppercase letters. Not
4742 all resolvers do this - some return always lower case. Exim was treating a
4743 change of case by a resolver as a change of domain, similar to a widening
4744 of a domain abbreviation. This triggered its re-routing code and so it was
4745 trying to route what was effectively the same domain again. This normally
4746 caused routing to fail (because the router wouldn't handle the domain
4747 twice). Now Exim checks for this case specially, and just changes the
4748 casing of the domain that it ultimately uses when it transmits the message
4751 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4754 50. If a filter generated a file delivery with a non-absolute name (possible if
4755 no home directory exists for the router), the forbid_file option was not
4758 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4759 the existing equality matching.
4761 52. Exim was using ints instead of ino_t variables in some places where it was
4762 dealing with inode numbers.
4764 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4765 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4766 called TMPDIR, and if it finds it is different, it changes its value.
4768 54. The smtp_printf() function is now made available to local_scan() so
4769 additional output lines can be written before returning. There is also an
4770 smtp_fflush() function to enable the detection of a dropped connection.
4771 The variables smtp_input and smtp_batched_input are exported to
4774 55. Changed the default runtime configuration: the message "Unknown user"
4775 has been removed from the ACL, and instead placed on the localuser router,
4776 using the cannot_route_message feature. This means that any verification
4777 failures that generate their own messages won't get overridden. Similarly,
4778 the "Unrouteable address" message that was in the ACL for unverifiable
4779 relay addresses has also been removed.
4781 56. Added hosts_avoid_esmtp to the smtp transport.
4783 57. The exicyclog script was not checking for the esoteric option
4784 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4785 will work only if exicyclog is run under the appropriate euid.
4787 58. Following a discussion on the list, the rules by which Exim recognises line
4788 endings on incoming messages have been changed. The -dropcr and drop_cr
4789 options are now no-ops, retained only for backwards compatibility. The
4790 following line terminators are recognized: LF CRLF CR. However, special
4791 processing applies to CR:
4793 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4794 nor a local message in the state where . is a terminator.
4796 (ii) If a bare CR is encountered in a header line, an extra space is added
4797 after the line terminator so as not to end the header. The reasoning
4798 behind this is that bare CRs in header lines are most likely either
4799 to be mistakes, or people trying to play silly games.
4801 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4802 was being incorrectly given as 18 bytes larger than it should have been.
4803 This is a VOB (very old bug).
4805 60. This may never have affected anything current, but just in case it has:
4806 When the local host is found other than at the start of a list of hosts,
4807 the local host, those with the same MX, and any that follow, are discarded.
4808 When the list in question was part of a longer list of hosts, the following
4809 hosts (not currently being processed) were also being discarded. This no
4810 longer happens. I'm not sure if this situation could ever has previously
4813 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4816 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4817 header. This is something that is recommended in a new Internet Draft, and
4818 is something that is documented as being done by Sendmail. There are two
4819 possible values. For messages generated by the autoreply transport, Exim
4822 Auto-Submitted: auto-replied
4824 whereas for all other generated messages (e.g. bounces) it adds
4826 Auto-Submitted: auto-generated
4828 63. The "personal" condition in filters now includes a test for the
4829 Auto-Submitted: header. If it contains the string "auto-" the message it
4830 not considered personal.
4832 64. Added rcpt_include_affixes as a generic transport option.
4834 65. Added queue_only_override (default true).
4836 66. Added the syslog_duplication option.
4838 67. If what should have been the first header line of a message consisted of
4839 a space followed by a colon, Exim was mis-interpreting it as a header line.
4840 It isn't of course - it is syntactically invalid and should therefore be
4841 treated as the start of the message body. The misbehaviour could have
4842 caused a number of strange effects, including loss of data in subsequent
4843 header lines, and spool format errors.
4845 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4846 client host had authenticated. This control can now be exercised by an ACL
4847 for more flexibility.
4849 69. By default, callouts do not happen when testing with -bh. There is now a
4850 variant, -bhc, which does actually run the callout code, including
4851 consulting and updating the callout cache.
4853 70. Added support for saslauthd authentication, courtesy of Alexander
4856 71. If statvfs() failed on the spool or log directories while checking their
4857 size for availability, Exim confusingly gave the error "space shortage".
4858 Furthermore, in debugging mode it crashed with a floating point exception.
4859 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4860 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4861 serious problem, Exim now writes to the main and panic logs when this
4862 happens, with details of the failure. It then refuses to accept the
4863 incoming message, giving the message "spool directory problem" or "log
4864 directory problem" with a 421 code for SMTP messages.
4866 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4867 0, 1, and 2 exist, because some OS complain for execs without them (see
4868 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4869 descriptors. However, the code omitted to check that the open succeeded,
4870 causing mysterious errors if for some reason the permissions on /dev/null
4871 got screwed. Now Exim writes a message to the main and panic logs, and
4872 bombs out if it can't open /dev/null.
4874 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4875 interact so that it is all more flexible. It is supposed to remain
4876 backwards compatible. Also added extra_local_interfaces.
4878 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4879 to bomb out with an assertion failure - to the client this appears as a
4880 connection drop. This problem occurs in the part of the code that was taken
4881 from the Samba project. Fortunately, the assertion is in a very simple
4882 function, so I have fixed this by reproducing the function inline in the
4883 one place where it is called, and arranging for authentication to fail
4884 instead of killing the process with assert().
4886 75. The SPA client code was not working when the server requested OEM rather
4887 than Unicode encoding.
4889 76. Added code to make require_files with a specific uid setting more usable in
4890 the case where statting the file as root fails - usually a non-root-mounted
4891 NFS file system. When this happens and the failure is EACCES, Exim now
4892 forks a subprocess and does the per-uid checking as the relevant uid.
4894 77. Added process_log_path.
4896 78. If log_file_path was not explicitly set, a setting of check_log_space or
4897 check_log_inodes was ignored.
4899 79. If a space check for the spool or log partitions fails, the incident is now
4900 logged. Of course, in the latter case the data may get lost...
4902 80. Added the %p formatting code to string_format() so that it can be used to
4903 print addresses in debug_print(). Adjusted all the address printing in the
4904 debugging in store.c to use %p rather than %d.
4906 81. There was a concern that a line of code in smtp_in.c could overflow a
4907 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4908 initially expressed, the concern was not well-founded, because trailing
4909 spaces are removed early. However, if the trailing spaces were followed by
4910 a NULL, they did not get removed, so the overflow was possible. Two fixes
4913 (a) I re-wrote the offending code in a cleaner fashion.
4914 (b) If an incoming SMTP command contains a NULL character, it is rejected
4917 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4918 runs initgroups(), so that if the Exim user is in any additional groups,
4919 they will be used during message reception.
4925 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob