1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.549 2008/07/18 17:55:42 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.
64 TK/01 Add preliminary DKIM support. Currently requires a forked version of
65 ALT-N's libdkim that I have put here:
66 http://duncanthrax.net/exim-experimental/
68 Note to Michael Haardt: I had to rename some vars in sieve.c. They
69 were called 'true' and it seems that C99 defines that as a reserved
70 keyword to be used with 'bool' variable types. That means you could
71 not include C99-style headers which use bools without triggering
72 build errors in sieve.c.
74 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
75 as mailq or other aliases. Changed the --help handling significantly
76 to do whats expected. exim_usage() emits usage/help information.
78 SC/01 Added the -bylocaldomain option to eximstats.
80 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
82 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
84 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
90 PH/01 Another patch from the Sieve maintainer.
92 PH/02 When an IPv6 address is converted to a string for single-key lookup
93 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
94 dots are used instead of colons so that keys in lsearch files need not
95 contain colons. This was done some time before quoting was made available
96 in lsearch files. However, iplsearch files do require colons in IPv6 keys
97 (notated using the quote facility) so as to distinguish them from IPv4
98 keys. This meant that lookups for IP addresses in host lists did not work
99 for iplsearch lookups.
101 This has been fixed by arranging for IPv6 addresses to be expressed with
102 colons if the lookup type is iplsearch. This is not incompatible, because
103 previously such lookups could never work.
105 The situation is now rather anomolous, since one *can* have colons in
106 ordinary lsearch keys. However, making the change in all cases is
107 incompatible and would probably break a number of configurations.
109 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
112 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
113 conversion specification without a maximum field width, thereby enabling
114 a rogue spamd server to cause a buffer overflow. While nobody in their
115 right mind would setup Exim to query an untrusted spamd server, an
116 attacker that gains access to a server running spamd could potentially
117 exploit this vulnerability to run arbitrary code as the Exim user.
119 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
120 $primary_hostname instead of what libspf2 thinks the hosts name is.
122 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
123 a directory entry by the name of the lookup key. Previously, if a
124 symlink pointed to a non-existing file or a file in a directory that
125 Exim lacked permissions to read, a lookup for a key matching that
126 symlink would fail. Now it is enough that a matching directory entry
127 exists, symlink or not. (Bugzilla 503.)
129 PH/03 The body_linecount and body_zerocount variables are now exported in the
132 PH/04 Added the $dnslist_matched variable.
134 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
135 This means they are set thereafter only if the connection becomes
138 PH/06 Added the client_condition to authenticators so that some can be skipped
139 by clients under certain conditions.
141 PH/07 The error message for a badly-placed control=no_multiline_responses left
142 "_responses" off the end of the name.
144 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
146 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
147 (without spaces) instead of just copying the configuration text.
149 PH/10 Added the /noupdate option to the ratelimit ACL condition.
151 PH/11 Added $max_received_linelength.
153 PH/12 Added +ignore_defer and +include_defer to host lists.
155 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
156 way in which PCRE > 7.0 is built.
158 PH/14 Implemented queue_only_load_latch.
160 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
161 MAIL command. The effect was to mangle the value on 64-bit systems.
163 PH/16 Another patch from the Sieve maintainer.
165 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
167 PH/18 If a system quota error occurred while trying to create the file for
168 a maildir delivery, the message "Mailbox is full" was not appended to the
169 bounce if the delivery eventually timed out. Change 4.67/27 below applied
170 only to a quota excession during the actual writing of the file.
172 PH/19 It seems that peer DN values may contain newlines (and other non-printing
173 characters?) which causes problems in log lines. The DN values are now
174 passed through string_printing() before being added to log lines.
176 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
177 and InterBase are left for another time.)
179 PH/21 Added message_body_newlines option.
181 PH/22 Guard against possible overflow in moan_check_errorcopy().
183 PH/23 POSIX allows open() to be a macro; guard against that.
185 PH/24 If the recipient of an error message contained an @ in the local part
186 (suitably quoted, of course), incorrect values were put in $domain and
187 $local_part during the evaluation of errors_copy.
193 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
194 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
197 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
198 issue a MAIL command.
200 PH/02 In an ACL statement such as
202 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
204 if a client was not listed at all, or was listed with a value other than
205 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
206 the condition was not true (as it should be), so access was not denied.
207 The bug was that the ! inversion was incorrectly passed on to the second
208 item. This has been fixed.
210 PH/03 Added additional dnslists conditions == and =& which are different from
211 = and & when the dns lookup returns more than one IP address.
213 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
214 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
216 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
217 FSYNC, which compiles an option called disable_fsync that allows for
218 bypassing fsync(). The documentation is heavily laced with warnings.
220 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
222 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
223 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
224 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
225 including adding "make clean"; (3) Added -fPIC when compiling the test
226 dynamically loaded module, to get rid of a warning.
228 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
229 message fails, move_frozen_messages = true and ignore_bounce_errors_after
230 = 0s. The bug is otherwise harmless.
232 PH/07 There was a bug in the dovecot authenticator such that the value of
233 $auth1 could be overwritten, and so not correctly preserved, after a
234 successful authentication. This usually meant that the value preserved by
235 the server_setid option was incorrect.
237 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
239 PH/09 Installed PCRE release 7.0.
241 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
242 run for batched SMTP input. It is now run at the start of every message
243 in the batch. While fixing this I discovered that the process information
244 (output by running exiwhat) was not always getting set for -bs and -bS
245 input. This is fixed, and it now also says "batched" for BSMTP.
247 PH/11 Added control=no_pipelining.
249 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
250 patch, slightly modified), and move the expansion of helo_data till after
251 the connection is made in the smtp transport (so it can use these
254 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
256 PH/14 Added log_selector = +pid.
258 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
260 PH/16 Add ${if forany and ${if forall.
262 PH/17 Added dsn_from option to vary the From: line in DSNs.
264 PH/18 Flush SMTP output before performing a callout, unless control =
265 no_callout_flush is set.
267 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
268 was true (the default) a successful delivery failed to delete the retry
269 item, thus causing premature timeout of the address. The bug is now
272 PH/20 Added hosts_avoid_pipelining to the smtp transport.
274 PH/21 Long custom messages for fakedefer and fakereject are now split up
275 into multiline reponses in the same way that messages for "deny" and
276 other ACL rejections are.
278 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
279 with slight modification.
281 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
282 draft, changing the syntax and factoring some duplicate code".
284 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
285 for deliveries of the second and subsequent messages over the same SMTP
288 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
289 ${reduce, with only minor "tidies".
291 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
293 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
294 expansion side effects.
296 PH/27 When a message times out after an over-quota error from an Exim-imposed
297 quota, the bounce message says "mailbox is full". This message was not
298 being given when it was a system quota that was exceeded. It now should
301 MH/03 Made $recipients available in local_scan(). local_scan() already has
302 better access to the recipient list through recipients_list[], but
303 $recipients can be useful in postmaster-provided expansion strings.
305 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
306 in the case of a MAIL command with additional options following the
307 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
308 were accidentally chopped off.
310 PH/29 SMTP synchronization checks are implemented when a command is read -
311 there is a check that no more input is waiting when there shouldn't be
312 any. However, for some commands, a delay in an ACL can mean that it is
313 some time before the response is written. In this time, more input might
314 arrive, invalidly. So now there are extra checks after an ACL has run for
315 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
316 pipelining has not been advertised.
318 PH/30 MH's patch to allow iscntrl() characters to be list separators.
320 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
321 returned in the SMTP response when smtp_return_error_details was false.
324 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
325 instead of the C I/O that was originally supplied, because problems were
328 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
329 Exim which did not show up earlier: it was assuming that a call to
330 SSL_CTX_set_info_callback() might give an error value. In fact, there is
331 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
332 was a macro that became an assignment, so it seemed to work. This has
333 changed to a proper function call with a void return, hence the compile
334 error. Exim's code has been fixed.
336 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
339 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
341 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
343 PH/37 If a message is not accepted after it has had an id assigned (e.g.
344 because it turns out to be too big or there is a timeout) there is no
345 "Completed" line in the log. When some messages of this type were
346 selected by exigrep, they were listed as "not completed". Others were
347 picked up by some special patterns. I have improved the selection
348 criteria to be more general.
350 PH/38 The host_find_failed option in the manualroute router can now be set
351 to "ignore", to completely ignore a host whose IP address cannot be
352 found. If all hosts are ignored, the behaviour is controlled by the new
353 host_all_ignored option.
355 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
356 homing or because of multiple MX records with /mx) generated more than
357 one IP address, and the following item turned out to be the local host,
358 all the secondary addresses of the first item were incorrectly removed
359 from the list, along with the local host and any following hosts (which
360 is what is supposed to happen).
362 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
363 whoever called Exim into the -H file. In the case of the daemon it was
364 behaving confusingly. When first started, it used values for whoever
365 started the daemon, but after a SIGHUP it used the Exim user (because it
366 calls itself on a restart). I have changed the code so that it now always
369 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
370 message are rejected with the same error (e.g. no authentication or bad
371 sender address), and a DATA command is nevertheless sent (as can happen
372 with PIPELINING or a stupid MUA), the error message that was given to the
373 RCPT commands is included in the rejection of the DATA command. This is
374 intended to be helpful for MUAs that show only the final error to their
377 PH/42 Another patch from the Sieve maintainer.
379 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
380 Eximstats - Fixed some broken HTML links and added missing column headers
382 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
383 columns for Rejects, Temp Rejects, Ham, and Spam rows.
385 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
387 PH/43 Yet another patch from the Sieve maintainer.
389 PH/44 I found a way to check for a TCP/IP connection going away before sending
390 the response to the final '.' that terminates a message, but only in the
391 case where the client has not sent further data following the '.'
392 (unfortunately, this is allowed). However, in many cases there won't be
393 any further data because there won't be any more messages to send. A call
394 to select() can be used: if it shows that the input is "ready", there is
395 either input waiting, or the socket has been closed. An attempt to read
396 the next input character can distinguish the two cases. Previously, Exim
397 would have sent an OK response which the client would never have see.
398 This could lead to message repetition. This fix should cure that, at
399 least in a lot of common cases.
401 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
402 advertised in response to EHLO.
408 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
409 fixed by 4.65/MH/01 (is this a record?) are fixed:
411 (i) An empty string was always treated as zero by the numeric comparison
412 operators. This behaviour has been restored.
414 (ii) It is documented that the numeric comparison operators always treat
415 their arguments as decimal numbers. This was broken in that numbers
416 starting with 0 were being interpreted as octal.
418 While fixing these problems I realized that there was another issue that
419 hadn't been noticed. Values of message_size_limit (both the global option
420 and the transport option) were treated as octal if they started with 0.
421 The documentation was vague. These values are now always treated as
422 decimal, and I will make that clear in the documentation.
428 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
429 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
432 MH/01 Don't check that the operands of numeric comparison operators are
433 integers when their expansion is in "skipping" mode (fixes bug
434 introduced by 4.64-PH/07).
436 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
437 child addresses, Exim now panics and dies. Previously, because the count
438 is held in a short int, deliveries were likely to be lost. As such a
439 large number of recipients for a single message is ridiculous
440 (performance will be very, very poor), I have chosen to impose a limit
441 rather than extend the field.
447 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
448 leftover -K file (the existence of which was triggered by #402).
449 While we were at it, introduced process PID as part of the -K
450 filename. This should rule out race conditions when creating
453 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
454 processing considerably. Previous code took too long for large mails,
455 triggering a timeout which in turn triggers #401.
457 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
458 in the DK code in transports.c. sendfile() is not really portable,
459 hence the _LINUX specificness.
461 TF/01 In the add_headers option to the mail command in an Exim filter,
462 there was a bug that Exim would claim a syntax error in any
463 header after the first one which had an odd number of characters
466 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
467 callout verification, Exim cached a "reject" for the entire domain. This
468 is correct for most verifications, but it is not correct for a recipient
469 verification with use_sender or use_postmaster set, because in that case
470 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
471 case of MAIL FROM:<> rejection from other early rejections (e.g.
472 rejection of HELO). When verifying a recipient using a non-null MAIL
473 address, the cache is ignored if it shows MAIL FROM:<> rejection.
474 Whatever the result of the callout, the value of the domain cache is
475 left unchanged (for any other kind of callout, getting as far as trying
476 RCPT means that the domain itself is ok).
478 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
481 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
482 manifest itself as EPIPE rather than ECONNECT. When tidying away a
483 session, the daemon ignores ECONNECT errors and logs others; it now
484 ignores EPIPE as well.
486 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
487 (quoted-printable decoding).
489 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
490 later the small subsequent patch to fix an introduced bug.
492 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
494 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
496 PH/08 An error is now given if message_size_limit is specified negative.
498 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
499 to be given (somewhat) arbitrary names.
501 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
504 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
505 miscellaneous code fixes
507 PH/10 Added the log_reject_target ACL modifier to specify where to log
510 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
511 hostname. This is wrong, because it relates to the incoming message (and
512 probably the interface on which it is arriving) and not to the outgoing
513 callout (which could be using a different interface). This has been
514 changed to use the value of the helo_data option from the smtp transport
515 instead - this is what is used when a message is actually being sent. If
516 there is no remote transport (possible with a router that sets up host
517 addresses), $smtp_active_hostname is used.
519 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
520 tweaks were necessary in order to get it to work (see also 21 below):
521 (a) The code assumed that strncpy() returns a negative number on buffer
522 overflow, which isn't the case. Replaced with Exim's string_format()
524 (b) There were several signed/unsigned issues. I just did the minimum
525 hacking in of casts. There is scope for a larger refactoring.
526 (c) The code used strcasecmp() which is not a standard C function.
527 Replaced with Exim's strcmpic() function.
528 (d) The code set only $1; it now sets $auth1 as well.
529 (e) A simple test gave the error "authentication client didn't specify
530 service in request". It would seem that Dovecot has changed its
531 interface. Fortunately there's a specification; I followed it and
532 changed what the client sends and it appears to be working now.
534 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
537 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
538 address A is aliased to B and C, where B exists and C does not. Without
539 -v the output is "A verified" because verification stops after a
540 successful redirection if more than one address is generated. However,
541 with -v the child addresses are also verified. Exim was outputting "A
542 failed to verify" and then showing the successful verification for C,
543 with its parentage. It now outputs "B failed to verify", showing B's
544 parentage before showing the successful verification of C.
546 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
547 look up a TXT record in a specific list after matching in a combined
550 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
551 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
552 they consult the DNS. I had assumed they would set it the way they
553 wanted; and indeed my experiments on Linux seem to show that in some
554 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
555 To be on the safe side, however, I have now made the interface to
556 host_find_byname() similar to host_find_bydns(), with an argument
557 containing the DNS resolver options. The host_find_byname() function now
558 sets these options at its start, just as host_find_bydns() does. The smtp
559 transport options dns_qualify_single and dns_search_parents are passed to
560 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
561 of host_find_byname() use the default settings of RES_DEFNAMES
562 (qualify_single) but not RES_DNSRCH (search_parents).
564 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
565 spool_read_header() do less string testing, by means of a preliminary
566 switch on the second character of optional "-foo" lines. (This is
567 overdue, caused by the large number of possibilities that now exist.
568 Originally there were few.) While I was there, I also converted the
569 str(n)cmp tests so they don't re-test the leading "-" and the first
570 character, in the hope this might squeeze out yet more improvement.
572 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
573 flag allowing group syntax was set by the header_syntax check but not
574 turned off, possible causing trouble later; (2) The flag was not being
575 set at all for the header_verify test, causing "group"-style headers to
576 be rejected. I have now set it in this case, and also caused header_
577 verify to ignore an empty address taken from a group. While doing this, I
578 came across some other cases where the code for allowing group syntax
579 while scanning a header line wasn't quite right (mostly, not resetting
580 the flag correctly in the right place). These bugs could have caused
581 trouble for malformed header lines. I hope it is now all correct.
583 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
584 with the "reply" argument non-NULL. The code, however (which originally
585 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
586 but it didn't always do it. This confused somebody who was copying the
587 code for some other use. I have removed all the tests.
589 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
590 feature that was used to support insecure browsers during the U.S. crypto
591 embargo. It requires special client support, and Exim is probably the
592 only MTA that supported it -- and would never use it because real RSA is
593 always available. This code has been removed, because it had the bad
594 effect of slowing Exim down by computing (never used) parameters for the
595 RSA_EXPORT functionality.
597 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
598 authenticator to fail if there's a tab character in the incoming data
599 (there should never be unless someone is messing about, as it's supposed
600 to be base64-encoded). Also added, on Timo's advice, the "secured" option
601 if the connection is using TLS or if the remote IP is the same as the
602 local IP, and the "valid-client-cert option" if a client certificate has
605 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
606 authenticators. This can be used for authorization after authentication
607 succeeds. (In the case of plaintext, it servers for both authentication
610 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
611 if any retry times were supplied.
613 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
614 connection, where there is no client IP address to check. In this
615 situation, the verify now always succeeds.
617 PH/25 Applied John Jetmore's -Mset patch.
619 PH/26 Added -bem to be like -Mset, but loading a message from a file.
621 PH/27 In a string expansion for a processed (not raw) header when multiple
622 headers of the same name were present, leading whitespace was being
623 removed from all of them, but trailing whitespace was being removed only
624 from the last one. Now trailing whitespace is removed from each header
625 before concatenation. Completely empty headers in a concatenation (as
628 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
629 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
631 PH/29 [Removed. This was a change that I later backed out, and forgot to
632 correct the ChangeLog entry (that I had efficiently created) before
633 committing the later change.]
635 PH/30 Exim was sometimes attempting to deliver messages that had suffered
636 address errors (4xx response to RCPT) over the same connection as other
637 messages routed to the same hosts. Such deliveries are always "forced",
638 so retry times are not inspected. This resulted in far too many retries
639 for the affected addresses. The effect occurred only when there were more
640 hosts than the hosts_max_try setting in the smtp transport when it had
641 the 4xx errors. Those hosts that it had tried were not added to the list
642 of hosts for which the message was waiting, so if all were tried, there
643 was no problem. Two fixes have been applied:
645 (i) If there are any address or message errors in an SMTP delivery, none
646 of the hosts (tried or untried) are now added to the list of hosts
647 for which the message is waiting, so the message should not be a
648 candidate for sending over the same connection that was used for a
649 successful delivery of some other message. This seems entirely
650 reasonable: after all the message is NOT "waiting for some host".
651 This is so "obvious" that I'm not sure why it wasn't done
652 previously. Hope I haven't missed anything, but it can't do any
653 harm, as the worst effect is to miss an optimization.
655 (ii) If, despite (i), such a delivery is accidentally attempted, the
656 routing retry time is respected, so at least it doesn't keep
657 hammering the server.
659 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
660 in ${readsocket because some servers need this prod.
662 PH/32 Added some extra debug output when updating a wait-xxx database.
664 PH/33 The hint "could be header name not terminated by colon", which has been
665 given for certain expansion errors for a long time, was not being given
666 for the ${if def:h_colon_omitted{... case.
668 PH/34 The spec says: "With one important exception, whenever a domain list is
669 being scanned, $domain contains the subject domain." There was at least
670 one case where this was not true.
672 PH/35 The error "getsockname() failed: connection reset by peer" was being
673 written to the panic log as well as the main log, but it isn't really
674 panic-worthy as it just means the connection died rather early on. I have
675 removed the panic log writing for the ECONNRESET error when getsockname()
678 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
679 runs only) independently of the message's sender address. This meant
680 that, if the 4xx error was in fact related to the sender, a different
681 message to the same recipient with a different sender could confuse
682 things. In particualar, this can happen when sending to a greylisting
683 server, but other circumstances could also provoke similar problems.
684 I have changed the default so that the retry time for these errors is now
685 based a combination of the sender and recipient addresses. This change
686 can be overridden by setting address_retry_include_sender=false in the
689 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
690 remote server are returned as part of bounce messages. This was not
691 happening for LMTP over a pipe (the lmtp transport), but now it is the
692 same for both kinds of LMTP.
694 PH/38 Despite being documented as not happening, Exim was rewriting addresses
695 in header lines that were in fact CNAMEs. This is no longer the case.
697 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
698 and queue runs started by the daemon processed all messages. This has
699 been fixed so that -R and -S can now usefully be given with -q<time>.
701 PH/40 Import PCRE release 6.7 (fixes some bugs).
703 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
705 PH/42 Give an error if -q is specified more than once.
707 PH/43 Renamed the variables $interface_address and $interface_port as
708 $received_ip_address and $received_port, to make it clear that these
709 values apply to message reception, and not to the outgoing interface when
710 a message is delivered. (The old names remain recognized, of course.)
712 PH/44 There was no timeout on the connect() call when using a Unix domain
713 socket in the ${readsocket expansion. There now is.
715 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
716 be meaningful with "accept".
718 SC/01 Eximstats V1.43
719 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
721 SC/02 Eximstats V1.44
722 Use a glob alias rather than an array ref in the generated
723 parser. This improves both readability and performance.
725 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
726 Collect SpamAssassin and rejection statistics.
727 Don't display local sender or destination tables unless
728 there is data to show.
729 Added average volumes into the top table text output.
731 SC/04 Eximstats V1.46
732 Collect data on the number of addresses (recipients)
733 as well as the number of messages.
735 SC/05 Eximstats V1.47
736 Added 'Message too big' to the list of mail rejection
737 reasons (thanks to Marco Gaiarin).
739 SC/06 Eximstats V1.48
740 Mainlog lines which have GMT offsets and are too short to
741 have a flag are now skipped.
743 SC/07 Eximstats V1.49 (Alain Williams)
744 Added the -emptyok flag.
746 SC/08 Eximstats V1.50
747 Fixes for obtaining the IP address from reject messages.
749 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
750 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
751 whitesspace changes from 4.64-PH/27
753 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
756 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
757 are found, allow negative numbers in numeric criteria)
759 JJ/06 exipick.20061117.2, added new $message_body_missing variable
761 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
762 to match changes made in 4.64-PH/43
764 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
766 PH/47 Put in an explicit test for a DNS lookup of an address record where the
767 "domain" is actually an IP address, and force a failure. This locks out
768 those revolvers/nameservers that support "A-for-A" lookups, in
769 contravention of the specifications.
771 PH/48 When a host name was looked up from an IP address, and the subsequent
772 forward lookup of the name timed out, the host name was left in
773 $sender_host_name, contrary to the specification.
775 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
776 restricted to single-key lookups, Exim was not diagnosing an error if
777 * or *@ was used with a query-style lookup.
779 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
781 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
782 long ago, but noone interested enough thought of it. Let's just say that
783 the "1.1" means that there are some new functions that weren't there at
784 some point in the past.
786 PH/51 Error processing for expansion failure of helo_data from an smtp
787 transport during callout processing was broken.
789 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
790 tested/used via the -bh/-bhc/-bs options.
792 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
793 bug, fixed in subsequent PCRE releases).
795 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
796 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
798 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
804 SC/01 Use a glob alias rather than an array ref in eximstats generated
805 parser. This improves both readability and performance.
807 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
808 Don't display local sender or destination tables in eximstats unless
809 there is data to show.
810 Added average volumes into the eximstats top table text output.
812 SC/03 Collect data on the number of addresses (recipients) as well
813 as the number of messages in eximstats.
815 TF/01 Correct an error in the documentation for the redirect router. Exim
816 does (usually) call initgroups() when daemonizing.
818 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
819 with consistent privilege compared to when running as a daemon.
821 TF/03 Note in the spec that $authenticated_id is not set for local
822 submissions from trusted users.
824 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
825 Thanks to Dean Brooks <dean@iglou.com> for the patch.
827 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
828 by adding some example configuration directives to the default
829 configuration file. A little bit of work is required to uncomment the
830 directives and define how usernames and passwords are checked, but
831 there is now a framework to start from.
833 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
834 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
835 without this. I don't know how relevant this is to other LDAP libraries.
837 PH/02 Add the verb name to the "unknown ACL verb" error.
839 PH/03 Magnus Holmgren's patch for filter_prepend_home.
841 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
843 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
844 directory not expanded when it should be if an expanded home directory
845 was set for the address (which is overridden by the transport).
847 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
850 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
851 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
852 because it is too late at that time, and has no effect.
854 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
855 security issue with \' (bugzilla #107). I could not use the
856 PQescapeStringConn() function, because it needs a PGconn value as one of
859 PH/08 When testing addresses using -bt, indicate those final addresses that
860 are duplicates that would not cause an additional delivery. At least one
861 person was confused, thinking that -bt output corresponded to deliveries.
862 (Suppressing duplicates isn't a good idea as you lose the information
863 about possibly different redirections that led to the duplicates.)
865 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
866 systems where poll() doesn't work, in particular OS X.
868 PH/10 Added more information to debugging output for retry time not reached.
870 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
871 operations in malware.c.
873 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
876 PH/13 If write_rejectlog was set false when logging was sent to syslog with
877 syslog_duplication set false, log lines that would normally be written
878 both the the main log and to the reject log were not written to syslog at
881 PH/14 In the default configuration, change the use of "message" in ACL warn
882 statements to "add_header".
884 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
885 not followed by a command (e.g. "seen endif").
887 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
888 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
891 PH/17 Added extra conditions to the default value of delay_warning_condition
895 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
896 { match{$h_precedence:}{(?i)bulk|list|junk} } \
897 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
900 The Auto-Submitted: and various List- headers are standardised, whereas I
901 don't think Precedence: ever was.
903 PH/18 Refactored debugging code in route_finduser() to show more information,
904 in particular, the error code if getpwnam() issues one.
906 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
907 This is apparently needed in addition to the PH/07 change above to avoid
908 any possible encoding problems.
910 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
911 but not after initializing Perl.
913 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
914 output them only if debugging. By default they are written stderr,
915 apparently, which is not desirable.
917 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
920 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
923 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
925 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
926 authenticated or an ident call has been made. Suppress the default
927 values for $authenticated_id and $authenticated_sender (but permit -oMai
928 and -oMas) when testing with -bh.
930 PH/24 Re-jigged the order of the tests in the default configuration so that the
931 tests for valid domains and recipients precede the DNS black list and CSA
932 tests, on the grounds that those ones are more expensive.
934 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
935 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
936 This bug exists in every version of Exim that I still have, right back to
939 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
940 However, an attempt to turn on case-sensitivity in a regex key by
941 including (?-i) didn't work because the subject string was already
942 lowercased, and the effects were non-intuitive. It turns out that a
943 one-line patch can be used to allow (?-i) to work as expected.
949 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
950 other effects) broke the use of negated acl sub-conditions.
952 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
955 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
956 "Deny" causes Exim to reject the incoming connection with a 554 error.
957 Unfortunately, if there is a major crisis, such as a disk failure,
958 tcp-wrappers gives "deny", whereas what one would like would be some
959 kind of temporary error. A kludge has been added to help with this.
960 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
961 554 error is used if errno is still zero or contains ENOENT (which occurs
962 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
965 PH/03 Add -lutil to the default FreeBSD LIBS setting.
967 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
968 errors. Otherwise a message that provokes a temporary error (when other
969 messages do not) can cause a whole host to time out.
971 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
972 addresses were routed directly to files or pipes from a redirect router.
973 File deliveries just didn't batch; pipe deliveries might have suffered
976 PH/06 A failure to get a lock for a hints database would erroneously always say
977 "Failed to get write lock", even when it was really a read lock.
979 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
980 of 0600. This has been changed to use the value of the lockfile_mode
981 option (which defaults to 0600).
983 PH/08 Applied small patch from the Sieve maintainer.
985 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
986 folder from quota calculations, a direct delivery into this folder messed
987 up the contents of the maildirsize file. This was because the regex was
988 used only to exclude .Trash (or whatever) when the size of the mailbox
989 was calculated. There was no check that a delivery was happening into an
990 excluded directory. This bug has been fixed by ignoring all quota
991 processing for deliveries into excluded directories.
993 PH/10 Added the maildirfolder_create_regex option to appendfile.
999 PH/01 The code for finding all the local interface addresses on a FreeBSD
1000 system running IPv6 was broken. This may well have applied to all BSD
1001 systems, as well as to others that have similar system calls. The broken
1002 code found IPv4 interfaces correctly, but gave incorrect values for the
1003 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1004 that it would not match correctly against @[] and not recognize the IPv6
1007 PH/02 The ipliteral router was not recognizing addresses of the form user@
1008 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1010 PH/03 Added disable_ipv6.
1012 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1013 decoded form, because it is most often used to construct To: headers
1014 lines in autoreplies, and the decoded form may well be syntactically
1015 invalid. However, $reply_address has leading white space removed, and all
1016 newlines turned into spaces so that the autoreply transport does not
1019 PH/05 If group was specified without a user on a router, and no group or user
1020 was specified on a transport, the group from the router was ignored.
1022 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1023 for visible compile-time settings that can be used to change these
1024 numbers, for those that want even more. Backwards compatibility with old
1025 spool files has been maintained. However, going back to a previous Exim
1026 release will lost any variables that are in spool files.
1028 PH/07 Two small changes when running in the test harness: increase delay when
1029 passing a TCP/IP connection to a new process, in case the original
1030 process has to generate a bounce, and remove special handling of
1031 127.0.0.2 (sic), which is no longer necessary.
1033 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1034 be the same on different OS.
1036 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1039 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1040 whether --show-vars was specified or not
1042 JJ/02 exipick: Added support for new ACL variable spool format introduced
1045 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1046 syntactically invalid From: or Reply-to: line, and a filter used this to
1047 generate an autoreply, and therefore failed to obtain an address for the
1048 autoreply, Exim could try to deliver to a non-existent relative file
1049 name, causing unrelated and misleading errors. What now happens is that
1050 it logs this as a hard delivery error, but does not attempt to create a
1053 PH/11 The exinext utility has a -C option for testing purposes, but although
1054 the given file was scanned by exinext itself; it wasn't being passed on
1055 when Exim was called.
1057 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1058 an end-of-file indication when reading a command response.
1060 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1061 compiled. In many other places in Exim, IPv6 addresses are always
1062 recognized, so I have changed this. It also means that IPv4 domain
1063 literals of the form [IPV4:n.n.n.n] are now always recognized.
1065 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1066 used if the router is not running as root, for example, when verifying at
1067 ACL time, or when using -bh. The debugging output from this situation was
1068 non-existent - all you got was a failure to exec. I have made two
1071 (a) Failures to set uid/gid, the current directory, or a process leader
1072 in a subprocess such as that created by queryprogram now generate
1073 suitable debugging ouput when -d is set.
1075 (b) The queryprogram router detects when it is not running as root,
1076 outputs suitable debugging information if -d is set, and then runs
1077 the subprocess without attempting to change uid/gid.
1079 PH/15 Minor change to Makefile for building test_host (undocumented testing
1082 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1083 additional section of a DNS packet that returns MX or SRV records.
1084 Instead, it always explicitly searches for A/AAAA records. This avoids
1085 major problems that occur when a DNS server includes only records of one
1086 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1087 fixed another bug: if SRV records were looked up and the corresponding
1088 address records were *not* found in the additional section, the port
1089 values from the SRV records were lost.
1091 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1092 using the correct key (the original address) when searching the retry
1093 rules in order to find which one to use for generating the retry hint.
1095 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1096 adding the default one. Similarly, if it contains a Reply-To: header, the
1097 errors_reply_to option, if set, is not used.
1099 PH/19 When calculating a retry time, Exim used to measure the "time since
1100 failure" by looking at the "first failed" field in the retry record. Now
1101 it does not use this if it is later than than the arrival time of the
1102 message. Instead it uses the arrival time. This makes for better
1103 behaviour in cases where some deliveries succeed, thus re-setting the
1104 "first failed" field. An example is a quota failure for a huge message
1105 when small messages continue to be delivered. Without this change, the
1106 "time since failure" will always be short, possible causing more frequent
1107 delivery attempts for the huge message than are intended.
1108 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1110 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1111 $1, $2, $3) because the numerical variables can be reset during some
1112 expansion items (e.g. "match"), thereby losing the authentication data.
1114 PH/21 Make -bV show the size of off_t variables so that the test suite can
1115 decide whether to run tests for quotas > 2G.
1117 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1118 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1119 filecount value is greater than 2G or if a quota value is greater than 2G
1120 on a system where the size of off_t is not greater than 4, a panic error
1123 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1124 never match. The debug and -bh output now contains an explicit error
1125 message indicating a malformed IPv4 address or mask.
1127 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1128 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1129 PH/23 above applies.
1131 PH/25 Do not write to syslog when running in the test harness. The only
1132 occasion when this arises is a failure to open the main or panic logs
1133 (for which there is an explicit test).
1135 PH/26 Added the /no_tell option to "control=freeze".
1137 PH/27 If a host name lookup failed very early in a connection, for example, if
1138 the IP address matched host_lookup and the reverse lookup yielded a name
1139 that did not have a forward lookup, an error message of the form "no IP
1140 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1141 could be logged. Now it outputs the IP address instead of "NULL".
1143 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1144 allows the sender and the authenticated sender to be set when
1145 submitting a message from within Exim. Since child_open_exim() is
1146 documented for local_scan(), the new function should be too.
1148 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1149 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1150 results in an empty string is now treated as unset.
1152 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1154 PH/31 Added sender_verify_fail logging option.
1156 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1157 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1158 tidied the source and removed it altogether.
1160 PH/33 When a queue run was abandoned because the load average was too high, a
1161 log line was always written; now it is written only if the queue_run log
1162 selector is set. In addition, the log line for abandonment now contains
1163 information about the queue run such as the pid. This is always present
1164 in "start" and "stop" lines but was omitted from the "abandon" line.
1166 PH/34 Omit spaces between a header name and the colon in the error message that
1167 is given when verify = headers_syntax fails (if there are lots of them,
1168 the message gets confusing).
1170 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1171 names, as there are now some PTR records that contain slashes. This check
1172 is only to protect against broken name servers that fall over on strange
1173 characters, so the fact that it applies to all lookups doesn't matter.
1175 PH/36 Now that the new test suite is complete, we can remove some of the
1176 special code in Exim that was needed for the old test suite. For example,
1177 sorting DNS records because real resolvers return them in an arbitrary
1178 order. The new test suite's fake resolver always returns records in the
1181 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1182 bounces) except when queue_only is set, to avoid logging races between
1183 the different processes.
1185 PH/38 Panic-die if .include specifies a non-absolute path.
1187 PH/39 A tweak to the "H" retry rule from its user.
1189 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1190 a label. They prevented compilation on older perls.
1192 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1193 a warning to be raised on newish perls.
1195 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1196 on queue. Changes to match documented behaviour of showing count of
1197 messages matching specified criteria.
1199 PH/40 Changed the default ident timeout from 30s to 5s.
1201 PH/41 Added support for the use of login_cap features, on those BSD systems
1202 that have them, for controlling the resources used by pipe deliveries.
1204 PH/42 The content-scanning code uses fopen() to create files in which to put
1205 message data. Previously it was not paying any attention to the mode of
1206 the files. Exim runs with umask(0) because the rest of the code creates
1207 files with open(), and sets the required mode explicitly. Thus, these
1208 files were ending up world-writeable. This was not a big issue, because,
1209 being within the spool directory, they were not world-accessible. I have
1210 created a function called modefopen, which takes an additional mode
1211 argument. It sets umask(777), creates the file, chmods it to the required
1212 mode, then resets the umask. All the relevant calls to fopen() in the
1213 content scanning code have been changed to use this function.
1215 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1216 to 24 hours. This avoids potential overflow problems when processing G
1217 and H retry rules. I suspect nobody ever tinkers with this value.
1219 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1221 PH/45 When the plaintext authenticator is running as a client, the server's
1222 challenges are checked to ensure they are valid base64 strings. By
1223 default, the authentication attempt is cancelled if an invalid string is
1224 received. Setting client_ignore_invalid_base64 true ignores these errors.
1225 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1226 they are received. Thus, the responses can be made to depend on the
1227 challenges. If an invalid string is ignored, an empty string is placed in
1230 PH/46 Messages that are created by the autoreply transport now contains a
1231 References: header, in accordance with RFCs 2822 and 3834.
1233 PH/47 Added authenticated_sender_force to the smtp transport.
1235 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1237 PH/49 Installed latest patch from the Sieve maintainer.
1239 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1240 was also set, the appendfile transport was unnecessarily scanning a
1241 directory of message files (e.g. for maildir delivery) to find the count
1242 of files (along with the size), even though it did not need this
1243 information. It now does the scan only if it needs to find either the
1244 size of the count of files.
1246 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1248 PH/52 Two bugs concerned with error handling when the smtp transport is
1251 (i) Exim was not creating retry information for temporary errors given
1252 for individual recipients after the DATA command when the smtp transport
1253 was used in LMTP mode. This meant that they could be retried too
1254 frequently, and not timed out correctly.
1256 (ii) Exim was setting the flag that allows error details to be returned
1257 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1258 recipients that were returned after the DATA command.
1260 PH/53 This is related to PH/52, but is more general: for any failing address,
1261 when detailed error information was permitted to be returned to the
1262 sender, but the error was temporary, then after the final timeout, only
1263 "retry timeout exceeded" was returned. Now it returns the full error as
1264 well as "retry timeout exceeded".
1266 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1267 do this, and (what is worse) MTAs that accept it.
1269 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1270 will now be deprecated.
1272 PH/56 New os.c-cygwin from the Cygwin maintainer.
1274 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1275 formats (previously only available in exim formats via -bpr, -bpru,
1276 and -bpra. Now also available in native and exiqgrep formats)
1278 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1279 with very large, slow to parse queues
1281 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1283 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1285 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1286 responses to authentication challenges, though it was showing the
1287 challenges; (ii) I've removed the CR characters from the debug output for
1290 PH/58 Allow for the insertion of a newline as well as a space when a string
1291 is turned into more than one encoded-word during RFC 2047 encoding. The
1292 Sieve code now uses this.
1294 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1295 data_4xx, lost_connection, tls_required.
1297 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1298 message was being sent as an SMTP response.
1300 PH/61 Add -l and -k options to exicyclog.
1302 PH/62 When verifying, if an address was redirected to one new address, so that
1303 verification continued, and the new address failed or deferred after
1304 having set something in $address_data, the value of $address_data was not
1305 passed back to the ACL. This was different to the case when no
1306 redirection occurred. The value is now passed back in both cases.
1308 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1309 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1310 use login_cap.h, so on its own it isn't the distinguishing feature. The
1311 new name refers directly to the setclassresources() function.
1313 PH/65 Added configuration files for NetBSD3.
1315 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1317 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1318 is preferred over IPv4.
1320 PH/68 The bounce_return_message and bounce_return_body options were not being
1321 honoured for bounces generated during the reception of non-SMTP messages.
1322 In particular, this applied to messages rejected by the ACL. This bug has
1323 been fixed. However, if bounce_return_message is true and bounce_return_
1324 body is false, the headers that are returned for a non-SMTP message
1325 include only those that have been read before the error was detected.
1326 (In the case of an ACL rejection, they have all been read.)
1328 PH/69 The HTML version of the specification is now built in a directory called
1329 spec_html instead of spec.html, because the latter looks like a path with
1330 a MIME-type, and this confuses some software.
1332 PH/70 Catch two compiler warnings in sieve.c.
1334 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1335 function verify_get_ident() calls ip_connect() to connect a socket, but
1336 if the "connect()" function timed out, ip_connect() used to close the
1337 socket. However, verify_get_ident() also closes the socket later, and in
1338 between Exim writes to the log, which may get opened at this point. When
1339 the socket was closed in ip_connect(), the log could get the same file
1340 descriptor number as the socket. This naturally causes chaos. The fix is
1341 not to close the socket in ip_connect(); the socket should be closed by
1342 the function that creates it. There was only one place in the code where
1343 this was missing, in the iplookup router, which I don't think anybody now
1344 uses, but I've fixed it anyway.
1346 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1347 well as to direct DNS lookups. Otherwise the handling of names in host
1348 lists is inconsistent and therefore confusing.
1354 PH/01 Two changes to the default runtime configuration:
1356 (1) Move the checks for relay_from_hosts and authenticated clients from
1357 after to before the (commented out) DNS black list checks.
1359 (2) Add control=submission to the relay_from_hosts and authenticated
1360 clients checks, on the grounds that messages accepted by these
1361 statements are most likely to be submissions.
1363 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1365 (1) Generate an error if the third argument for the ${prvs expansion is
1368 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1371 (3) Reset the variables that are obtained from the first argument of
1372 ${prvscheck and used in the second argument before leaving the code,
1373 because their memory is reclaimed, so using them afterwards may do
1376 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1377 one (it's much easier than Tom thought :-).
1379 (5) Because of (4), we can now allow for the use of $prvscheck_result
1380 inside the third argument.
1382 PH/03 For some reason, the default setting of PATH when running a command from
1383 a pipe transport was just "/usr/bin". I have changed it to
1386 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1387 anything to be listed in the output from -bV.
1389 PH/05 When a filter generated an autoreply, the entire To: header line was
1390 quoted in the delivery log line, like this:
1392 => >A.N.Other <ano@some.domain> <original@ddress> ...
1394 This has been changed so that it extracts the operative address. There
1395 may be more than one such address. If so, they are comma-separated, like
1398 => >ano@some.domain,ona@other.domain <original@ddress> ...
1400 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1401 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1402 not being looked up in the rDNS to get a host name, Exim was showing the
1403 IP address twice in Received: lines, even though the IP addresses were
1404 identical. For example:
1406 Received: from [1.2.3.4] (helo=[1.2.3.4])
1408 However, if the real host name was known, it was omitting the HELO data
1409 if it matched the actual IP address. This has been tidied up so that it
1410 doesn't show the same IP address twice.
1412 PH/07 When both +timestamp and +memory debugging was on, the value given by
1413 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1414 called by the debug printing, thereby overwriting the timestamp buffer.
1415 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1417 PH/08 When the original message was included in an autoreply transport, it
1418 always said "this is a copy of the message, including all the headers",
1419 even if body_only or headers_only was set. It now gives an appropriate
1422 PH/09 Applied a patch from the Sieve maintainer which:
1424 o fixes some comments
1425 o adds the (disabled) notify extension core
1426 o adds some debug output for the result of if/elsif tests
1427 o points to the current vacation draft in the documentation
1428 and documents the missing references header update
1432 o fixes a bug in processing the envelope test (when testing
1433 multiple envelope elements, the last element determinted the
1436 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1437 Electronic Mail") by including:
1439 Auto-submitted: auto-generated
1441 in the messages that it generates (bounce messages and others, such as
1442 warnings). In the case of bounce messages for non-SMTP mesages, there was
1443 also a typo: it was using "Auto_submitted" (underscore instead of
1444 hyphen). Since every message generated by Exim is necessarily in response
1445 to another message, thes have all been changed to:
1447 Auto-Submitted: auto-replied
1449 in accordance with these statements in the RFC:
1451 The auto-replied keyword:
1453 - SHOULD be used on messages sent in direct response to another
1454 message by an automatic process,
1456 - MUST NOT be used on manually-generated messages,
1458 - MAY be used on Delivery Status Notifications (DSNs) and Message
1459 Disposition Notifications (MDNs),
1461 - MUST NOT be used on messages generated by automatic or periodic
1462 processes, except for messages which are automatic responses to
1465 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1466 to the default Received: header definition.
1468 PH/12 Added log selector acl_warn_skipped (default on).
1470 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1471 variables because (a) they are in the wrong storage pool and (b) even if
1472 they were copied, it wouldn't work properly because of the caching.
1474 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1475 checking when decoding. Apparently there are clients that generate
1476 overlong encoded strings. Why am I not surprised?
1478 PH/15 If the first argument of "${if match_address" was not empty, but did not
1479 contain an "@" character, Exim crashed. Now it writes a panic log message
1480 and treats the condition as false.
1482 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1484 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1485 extension "envelope-auth". The code is finished and in agreement with
1486 other implementations, but there is no documentation so far and in fact,
1487 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1488 not changing the active code.
1490 Print executed "if" and "elsif" statements when debugging is used. This
1491 helps a great deal to understand what a filter does.
1493 Document more things not specified clearly in RFC3028. I had all this
1494 sorted out, when out of a sudden new issues came to my mind. Oops."
1496 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1499 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1500 canonical form (as documented). However, after a host name lookup from
1501 the IP address, check_host() was doing a simple string comparison with
1502 addresses acquired from the DNS when checking that the found name did
1503 have the original IP as one of its addresses. Since any found IPv6
1504 addresses are likely to be in abbreviated form, the comparison could
1505 fail. Luckily, there already exists a function for doing the comparison
1506 by converting both addresses to binary, so now that is used instead of
1507 the text comparison.
1509 PH/20 There was another similar case to PH/19, when a complete host name was
1510 given in a host list; looking up its IP address could give an abbreviated
1511 form, whereas the current host's name might or might not be abbreviated.
1512 The same fix has been applied.
1518 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1519 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1522 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1523 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1525 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1527 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1528 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1529 Also, for "Auto-Submitted", treat anything other than "no" as
1530 non-personal, in accordance with RFC 3834. (Previously it treated
1531 anything starting "auto-" as non-personal.)
1533 TF/01 The control=submission/name=... option had a problem with syntax
1534 errors if the name included a slash character. The /name= option
1535 now slurps the rest of the string, so it can include any characters
1536 but it must come last in the list of options (after /sender_retain
1539 PH/05 Some modifications to the interface to the fake nameserver for the new
1547 TK/01 Added the "success_on_redirect" address verification option. See
1548 NewStuff for rationale and an example.
1550 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1552 PH/02 Patch to exigrep to allow it to work on syslog lines.
1554 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1555 fread() to skip over the body file's header line, because in Cygwin the
1556 header line is locked and is inaccessible.
1558 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1559 co-exist for some time) to make it clear that it is the Exim ID that is
1560 referenced, not the Message-ID: header line.
1562 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1563 string_format() function, because snprintf() does not exist on all
1566 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1567 ${stat: expansion item.
1569 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1570 protocol synchronization error", to keep the pedants happy.
1572 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1573 well as for IRIX systems, when gcc is being used. See the host.c source
1576 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1578 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1581 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1582 transport and to the smtp transport in LMTP mode.
1584 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1586 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1587 run a filter in a subprocess. This could lead to confusion in subsequent
1588 lookups in the parent process. There should also be a search_tidyup() at
1589 the end of the subprocess.
1591 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1592 only if the host matched helo_try_verify_hosts, which caused the
1593 verification to occur when the EHLO/HELO command was issued. The ACL just
1594 tested the remembered result. Now, if a previous verification attempt has
1595 not happened, "verify = helo" does it there and then.
1597 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1599 TK/03 Fix log output including CR from clamd.
1601 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1602 exist provoked a memory error which could cause a segfault.
1604 PH/15 Installed PCRE 6.2
1606 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1608 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1609 of the problem. Specifically, suggested +O2 rather than +O1 for the
1612 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1614 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1615 record, the port was not correctly logged when the outgoing_port log
1616 selector was set (it logged the transort's default port).
1618 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1619 fallback_hosts, and "hosts" in the smtp transport.
1621 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1622 host errors such as "Connection refused".
1624 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1625 authentication with radiusclient 0.4.9:
1627 - Error returned from rc_read_config was caught wrongly
1628 - Username/password not passed on to radius server due to wrong length.
1630 The presumption is that some radiusclient API changes for 4.51/PH/17
1631 were not taken care of correctly. The code is still untested by me (my
1632 Linux distribution still has 0.3.2 of radiusclient), but it was
1633 contributed by a Radius user.
1635 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1636 expanding the "port" option of the smtp transport.
1638 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1639 while reading a MIME header. Thanks to Tom Hughes for a patch.
1641 PH/24 Include config.h inside local_scan.h so that configuration settings are
1644 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1645 that in an ACL for RCPT (for example), you can examine exactly what was
1648 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1649 commands, but it was not correctly comparing the address with the actual
1650 client host address. Thus, it would show the EHLO address in Received:
1651 header lines when this was not necessary.
1653 PH/27 Added the % operator to ${eval:}.
1655 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1656 it should be ignoring failures (because with -C, for example, it has lost
1657 privilege). It wasn't ignoring creation failures other than "already
1660 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1663 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1664 because an input line was too long, either on its own, or by virtue of
1665 too many continuations, the temporary file was not being removed, and the
1666 return code was incorrect.
1668 PH/31 Missing "BOOL" in function definition in filtertest.c.
1670 PH/32 Applied Sieve patches from the maintainer.
1672 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1674 PH/33 Added "verify = not_blind".
1676 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1677 Local/Makefile (with some defaults set). These are used in built scripts
1678 such as exicyclog, but they have never been used in the exim_install
1679 script (though there are many overriding facilities there). I have
1680 arranged that the exim_install script now takes note of these two
1683 PH/35 Installed configuration files for Dragonfly.
1685 PH/36 When a locally submitted message by a trusted user did not contain a
1686 From: header, and the sender address was obtained from -f or from an SMTP
1687 MAIL command, and the trusted user did not use -F to supply a sender
1688 name, $originator_name was incorrectly used when constructing a From:
1689 header. Furthermore, $originator_name was used for submission mode
1690 messages from external hosts without From: headers in a similar way,
1691 which is clearly wrong.
1693 PH/37 Added control=suppress_local_fixups.
1695 PH/38 When log_selector = +received_sender was set, and the addition of the
1696 sender made the log line's construction buffer exactly full, or one byte
1697 less than full, an overflow happened when the terminating "\n" was
1700 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1701 when the result of a list match is failure because a DNS lookup failed.
1703 PH/40 RM_COMMAND is now used in the building process.
1705 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1706 the "build-* directories that it finds.
1708 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1709 address in a domain literal was a prefix of an interface address.
1711 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1712 when verifying a sender address, unless rewrite_headers is false.
1714 PH/44 Wrote a long comment about why errors_to addresses are verified as
1715 recipients, not senders.
1717 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1718 the ratelimit ACL was added.
1720 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1722 PH/46 Added extra information about PostgreSQL errors to the error string.
1724 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1725 avoiding the need to install special zones in a real server. This is
1726 backwards compatible; if it can't find the fake resolver, it drops back.
1727 Thus, both old and new test suites can be run.
1729 TF/02 Added util/ratelimit.pl
1731 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1732 clock is set back in time.
1734 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1735 Candler <B.Candler@pobox.com>.
1737 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1738 OK for addresses that are the result of redirections.
1740 PH/48 A number of further additions for the benefit of the new test suite,
1741 including a fake gethostbyname() that interfaces to the fake DNS resolver
1744 TF/06 The fix for widen_domains has also been applied to qualify_single and
1745 search_parents which are the other dnslookup options that can cause
1748 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1751 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1753 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1754 service" if it encountered an MX record with an empty target hostname.
1755 The message is now "an MX or SRV record indicated no SMTP service".
1757 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1758 if the DNS of the sending site is misconfigured. This is quite a
1759 common situation. This change restores the behaviour of treating a
1760 helo verification defer as a failure.
1762 PH/51 If self=fail was set on a router, the bounce message did not include the
1763 actual error message.
1769 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1771 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1772 command itself ended in error, the underlying message about the transport
1773 filter timeout was being overwritten with the pipe command error. Now the
1774 underlying error message should be appended to the second error message.
1776 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1778 PH/02 Reduce the amount of output that "make" produces by default. Full output
1779 can still be requested.
1781 PH/03 The warning log line about a condition test deferring for a "warn" verb
1782 was being output only once per connection, rather than after each
1783 occurrence (because it was using the same function as for successful
1784 "warn" verbs). This seems wrong, so I have changed it.
1786 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1787 it should not have, which might have caused a crash in the right
1788 circumstances, but probably never did.
1790 PH/04 Installed a modified version of Tony Finch's patch to make submission
1791 mode fix the return path as well as the Sender: header line, and to
1792 add a /name= option so that you can make the user's friendly name appear
1795 TF/03 Added the control = fakedefer ACL modifier.
1797 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1798 Mark Lowes for thorough testing.
1800 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1802 TK/03 Merged latest SRS patch from Miles Wilton.
1804 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1805 with the definition in sysexits.h (which is #included earlier).
1806 Fortunately, Exim does not actually use EX_OK. The code used to try to
1807 preserve the sysexits.h value, by assumimg that macro definitions were
1808 scanned for macro replacements. I have been disabused of this notion,
1809 so now the code just undefines EX_OK before #including unistd.h.
1811 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1812 in the smtp transport. When a block could not be written in a single
1813 write() function, the timeout was being re-applied to each part-write.
1814 This seems wrong - if the receiver was accepting one byte at a time it
1815 would take for ever. The timeout is now adjusted when this happens. It
1816 doesn't have to be particularly precise.
1818 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1819 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1821 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1822 without a domain if the check to <postmaster@domain> fails.
1824 SC/01 Eximstats: added -xls and the ability to specify output files
1825 (patch written by Frank Heydlauf).
1827 SC/02 Eximstats: use FileHandles for outputing results.
1829 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1831 SC/04 Eximstats: fixed display of large numbers with -nvr option
1833 SC/05 Eximstats: fixed merging of reports with empty tables.
1835 SC/06 Eximstats: added the -include_original_destination flag
1837 SC/07 Eximstats: removed tabs and trailing whitespace.
1839 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1841 TK/06 MBOX spool code: Add real "From " MBOX separator line
1842 so the .eml file is really in mbox format (even though
1843 most programs do not really care). Patch from Alex Miller.
1845 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1846 The latter is generated from $received_to and is only set if the
1847 message has one envelope recipient. SA can use these headers,
1848 obviously out-of-the-box. Patch from Alex Miller.
1850 PH/08 The ${def test on a variable was returning false if the variable's
1851 value was "0", contrary to what the specification has always said!
1852 The result should be true unless the variable is empty.
1854 PH/09 The syntax error of a character other than { following "${if
1855 def:variable_name" (after optional whitespace) was not being diagnosed.
1856 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1857 accidental colon was present, for example, could give incorrect results.
1859 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1860 result is used (not including appendfile, where other changes are about
1863 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1864 This involved changing a lot of size variables from int to off_t. It
1865 should work with maildirs and everything.
1867 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1868 spamd dying while we are connected to it.
1870 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1873 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1874 function for BDB 4.3.
1876 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1878 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1879 the sources for PCRE have been re-arranged and more files are now
1882 PH/15 The code I had for printing potentially long long variables in PH/11
1883 above was not the best (it lost precision). The length of off_t variables
1884 is now inspected at build time, and an appropriate printing format (%ld
1885 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1886 to be "long long int" or "long int". This is needed for the internal
1887 formatting function string_vformat().
1889 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1890 the configuration file to be ":syslog", then the script "guesses" where
1891 the logs files are, rather than using the compiled in default. In our
1892 case the guess is not the same as the compiled default, so the script
1893 suddenly stopped working when I started to use syslog. The patch checks
1894 to see if log_file_path is "". If so, it attempts to read it from exim
1895 with no configuration file to get the compiled in version, before it
1896 falls back to the previous guessing code."
1898 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1899 implementing BATV in an Exim configuration. See NewStuff for the gory
1902 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1903 Makefile that are specific to HP-UX.
1905 PH/18 If the "use_postmaster" option was set for a recipient callout together
1906 with the "random" option, the postmaster address was used as the MAIL
1907 FROM address for the random test, but not for the subsequent recipient
1908 test. It is now used for both.
1910 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1911 patch removes a few documentation additions to RFC 3028, because the
1912 latest draft now contains them. It adds the new en;ascii-case comparator
1913 and a new error check for 8bit text in MIME parts. Comparator and
1914 require names are now matched exactly. I enabled the subaddress
1915 extension, but it is not well tested yet (read: it works for me)."
1917 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1918 rework some of the code of TK/09 above to avoid the hardwired use of
1919 "%lld" and "long long". Replaced the call to snprintf() with a call to
1922 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1923 records point to non-existent hosts", "retry timeout exceeded", and
1924 "retry time not reached for any host after a long failure period".
1926 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1927 experimental DomainKeys support:
1929 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1930 (2) On an error such as an illegally used "control", the wrong name for
1931 the control was given.
1933 These problems did NOT occur unless DomainKeys support was compiled.
1935 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1937 PH/24 Added ${if match_ip condition.
1939 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1940 databases so that it will be absolutely obvious if a crash occurs in the
1941 DB library. This is a regular occurrence (often caused by mis-matched
1944 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1945 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1946 picked up on a user's system that detects such things. There doesn't seem
1947 to be a gcc warning option for this - only an attribute that has to be
1948 put on the function's prototype. It seems that in Fedora Core 4 they have
1949 set this on a number of new functions. No doubt there will be more in due
1952 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1953 specify a transport. However, if an address that was verified by such a
1954 router was the subject of a callout, Exim crashed because it tried to
1955 read the rcpt_include_affixes from the non-existent transport. Now it
1956 just assumes that the setting of that option is false. This bug was
1957 introduced by 4.51/PH/31.
1959 PH/28 Changed -d+all to exclude +memory, because that information is very
1960 rarely of interest, but it makes the output a lot bigger. People tend to
1961 do -d+all out of habit.
1963 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1964 code in os-type was giving problems when libc.so lives in lib64, like on
1967 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1968 aren't the modern standard, and it seems that some systems' include files
1969 don't always have them. Exim was already checking for some of the newer
1970 ones like T_AAAA, and defining it itself. I've added checks for all the
1971 record types that Exim uses.
1973 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1974 not automatically generating a new one, as it is supposed to. This
1975 prevented TLS from working. If the file did exist, but contained invalid
1976 data, a new version was generated, as expected. It was only the case of a
1977 non-existent file that was broken.
1979 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1980 with a change in libdomainkeys > 0.64.
1982 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1983 from DNS. If the selector record carries the flag, it now has
1984 precedence over the domain-wide flag.
1986 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1988 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1989 the use of a port name that isn't defined in /etc/services) occurred, the
1990 message was deferred as in a normal delivery, and thus remained on the
1991 spool, instead of being failed because of the mua_wrapper setting. This
1992 is now fixed, and I tidied up some of the mua_wrapper messages at the
1995 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1996 the messages in a hash of arrays rather than using individual hashes.
1997 This is a bit cleaner and results in dramatic memory savings, albeit
1998 at a slight CPU cost.
2000 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2001 as requested by Marc Sherman.
2003 SC/10 Eximstats: added histograms for user specified patterns as requested
2006 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2008 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2009 fopen() in the content-scanning modules that did not already have it.
2015 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2016 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2018 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2020 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2022 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2023 to test Sieve filters that use "vacation".
2025 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2026 that changes the way the GnuTLS parameters are stored in the cache file.
2027 The new format can be generated externally. For backward compatibility,
2028 if the data in the cache doesn't make sense, Exim assumes it has read an
2029 old-format file, and it generates new data and writes a new file. This
2030 means that you can't go back to an older release without removing the
2033 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2034 work if there are any delivery delays because "one_time" forces the
2035 parent to be marked "delivered", so its unseen clone is never tried
2036 again. For this reason, Exim now forbids the simultaneous setting of
2039 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2040 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2041 entry if you want to know the details. The fix, however, neglected to
2042 consider the case when local delivery batching is involved. The test for
2043 "previously delivered" was not happening when checking to see if an
2044 address could be batched with a previous (undelivered) one; under
2045 certain circumstances this could lead to multiple deliveries to the same
2048 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2049 in its include files, and this causes problems building Exim.
2051 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2052 header_syntax) but Exim was just ignoring anything given after a slash.
2053 In particular, this caused confusion with an attempt to use "verify =
2054 reverse_host_lookup/defer_ok". An error is now given when options are
2055 supplied for verify items that do not have them. (Maybe reverse_host_
2056 lookup should have a defer_ok option, but that's a different point.)
2058 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2059 defined by RFC 821) to 2048, because there were problems with some AUTH
2060 commands, and RFC 1869 says the size should be increased for extended
2061 SMTP commands that take arguments.
2063 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2066 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2067 "unknown" error; now it says that the functionality isn't in the binary.
2069 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2070 an address' error message when a string expansion fails (syntax or
2071 whatever). Otherwise the password may appear in the log. Following change
2072 PH/42 below, there is no longer a chance of it appearing in a bounce
2075 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2077 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2078 of its addresses was ever tried. (Bugzilla bug #2.)
2080 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2081 the result incorrectly in the debug output. (It correctly added a newline
2082 to what was transported.)
2084 TF/01 Added $received_time.
2086 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2087 commented out examples of how to interface to a virus scanner and to
2088 SpamAssassin. Also added commented examples of av_scanner and
2089 spamd_address settings.
2091 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2092 and controls are allowed in which ACLs. There were a couple of minor
2093 errors. Some of the entries in the conditions table (which is a table of
2094 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2095 negation of where the condition IS allowed.
2097 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2099 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2100 header file does not have a version number, so I've had to invent a new
2101 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2102 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2103 radiusclient), but it was contributed by a Radius user.
2105 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2106 files or directories, for OpenSSL.
2108 PH/19 When an Exim process that is running as root has to create an Exim log
2109 file, it does so in a subprocess that runs as exim:exim so as to get the
2110 ownership right at creation (otherwise, other Exim processes might see
2111 the file with the wrong ownership). There was no test for failure of this
2112 fork() call, which would lead to the process getting stuck as it waited
2113 for a non-existent subprocess. Forks do occasionally fail when resources
2114 run out. I reviewed all the other calls to fork(); they all seem to check
2117 PH/20 When checking for unexpected SMTP input at connect time (before writing
2118 the banner), Exim was not dealing correctly with a non-positive return
2119 from the read() function. If the client had disconnected by this time,
2120 the result was a log entry for a synchronization error with an empty
2121 string after "input=" when read() returned zero. If read() returned -1
2122 (an event I could not check), uninitialized data bytes were printed.
2123 There were reports of junk text (parts of files, etc) appearing after
2126 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2128 PH/22 Added support for macro redefinition, and (re)definition in between
2129 driver and ACL definitions.
2131 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2132 forgetting to use the resulting value; it was using the unexpanded value.
2134 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2135 hadn't been configured. The fix is from Juergen Kreileder, who
2136 understands it better than I do:
2138 "Here's what I see happening with three configured cyrus_sasl
2139 authenticators configured (plain, login, cram-md5):
2141 On startup auth_cyrus_sasl_init() gets called for each of these.
2142 This means three calls to sasl_listmech() without a specified mech_list.
2143 => SASL tests which mechs of all available mechs actually work
2144 => three warnings about OTP not working
2145 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2147 With the patch, sasl_listmech() also gets called three times. But now
2148 SASL's mech_list option is set to the server_mech specified in the the
2149 authenticator. Or in other words, the answer from sasl_listmech()
2150 gets limited to just the mech you're testing for (which is different
2152 => the return list contains just 'plain' or 'login', 'cram-md5' or
2153 nothing depending on the value of ob->server_mech.
2155 I've just tested the patch: Authentication still works fine,
2156 unavailable mechs specified in the exim configuration are still
2157 caught, and the auth.log warnings about OTP are gone."
2159 PH/25 When debugging is enabled, the contents of the command line are added
2160 to the debugging output, even when log_selector=+arguments is not
2163 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2164 answer is "GNU", and only if the return is "GNU/something" is the answer
2167 PH/27 $acl_verify_message is now set immediately after the failure of a
2168 verification in an ACL, and so is available in subsequent modifiers. In
2169 particular, the message can be preserved by coding like this:
2171 warn !verify = sender
2172 set acl_m0 = $acl_verify_message
2174 Previously, $acl_verify_message was set only while expanding "message"
2175 and "log_message" when a very denied access.
2177 PH/28 Modified OS/os.c-Linux with
2179 -#ifndef OS_LOAD_AVERAGE
2180 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2182 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2183 nomenclature these days.)
2185 PH/29 Installed patch from the Sieve maintainer that adds the options
2186 sieve_useraddress and sieve_subaddress to the redirect router.
2188 PH/30 In these circumstances:
2189 . Two addresses routed to the same list of hosts;
2190 . First host does not offer TLS;
2191 . First host accepts first address;
2192 . First host gives temporary error to second address;
2193 . Second host offers TLS and a TLS session is established;
2194 . Second host accepts second address.
2195 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2196 and peerdn, if requested) that were in fact used only for the second
2199 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2200 attention to any local part prefix or suffix that was matched by the
2201 router that accepted the address. It now behaves in the same way as it
2202 does for delivery: the affixes are removed from the local part unless
2203 rcpt_include_affixes is set on the transport.
2205 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2206 timeout during the DATA phase of an incoming message.
2208 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2209 applied a patch sanctioned by the Sieve maintainer.
2211 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2212 the uid or gid is negative. A case of a negative gid caused this to be
2213 noticed. The fix allows for either to be negative.
2215 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2216 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2219 PH/36 The change PH/12 above was broken. Fixed it.
2221 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2222 the grounds that routing the same address twice would always produce the
2223 same answer. This might have been true once, but it is certainly no
2224 longer true now. Routing a child address may depend on the previous
2225 routing that produced that child. Some complicated redirection strategies
2226 went wrong when messages had multiple recipients, and made Exim's
2227 behaviour dependent on the order in which the addresses were given.
2229 I have moved the duplicate checking until after the routing is complete.
2230 Exim scans the addresses that are assigned to local and remote
2231 transports, and removes any duplicates. This means that more work will be
2232 done, as duplicates will always all be routed, but duplicates are
2233 presumably rare, so I don't expect this is of any significance.
2235 For deliveries to pipes, files, and autoreplies, the duplicate checking
2236 still happens during the routing process, since they are not going to be
2239 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2240 It corrects a timeout issue with spamd. This is Ian's comment: "The
2241 background is that sometimes spamd either never reads data from a
2242 connection it has accepted, or it never writes response data. The exiscan
2243 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2244 blindly assumes that writes won't block so it may never time out."
2246 PH/39 Allow G after quota size as well as K and M.
2248 PH/40 The value set for $authenticated_id in an authenticator may not contain
2249 binary zeroes or newlines because the value is written to log lines and
2250 to spool files. There was no check on this. Now the value is run through
2251 the string_printing() function so that such characters are converted to
2252 printable escape sequences.
2254 PH/41 $message_linecount is a new variable that contains the total number of
2255 lines in the message. Compare $body_linecount, which is the count for the
2258 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2259 bounce and delay warning messages, except in certain special cases, which
2262 (a) An SMTP error message from a remote host;
2263 (b) A message specified in a :fail: redirection;
2264 (c) A message specified in a "fail" command in a system filter;
2265 (d) A message specified in a FAIL return from the queryprogram router;
2266 (e) A message specified by the cannot_route_message router option.
2268 In these cases only, Exim does include the error details in bounce and
2269 warning messages. There are also a few cases where bland messages such
2270 as "unrouteable address" or "local delivery error" are given.
2272 PH/43 $value is now also set for the "else" part of a ${run expansion.
2274 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2275 being worked on, but at least Exim now implements the latest version to
2278 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2279 process to complete was treated as a delivery failure, a timeout while
2280 writing the message to the pipe was logged, but erroneously treated as a
2281 successful delivery. Such timeouts include transport filter timeouts. For
2282 consistency with the overall process timeout, these timeouts are now
2283 treated as errors, giving rise to delivery failures by default. However,
2284 there is now a new Boolean option for the pipe transport called
2285 timeout_defer, which, if set TRUE, converts the failures into defers for
2286 both kinds of timeout. A transport filter timeout is now identified in
2289 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2290 systems where "make" and "gmake" are different, calling "gmake" at top
2291 level broke things. I've arranged for the value of $(MAKE) to be passed
2292 from the Makefile to this script so that it can call the same version of
2296 A note about Exim versions 4.44 and 4.50
2297 ----------------------------------------
2299 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2300 changes of various kinds. As a consequence, a big documentation update was
2301 needed. This delayed the release for rather longer than seemed good, especially
2302 in the light of a couple of (minor) security issues. Therefore, the changes
2303 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2304 release. So 4.44 and 4.50 are in effect two different branches that both start
2307 I have left the 4.50 change log unchanged; it contains all the changes since
2308 4.43. The change log for 4.44 is below; many of its items are identical to
2309 those for 4.50. This seems to be the most sensible way to preserve the
2310 historical information.
2316 1. Minor wording change to the doc/README.SIEVE file.
2318 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2319 computation of the current number of files was incorrect.
2321 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2322 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2325 4. Give more explanation in the error message when the command for a transport
2326 filter fails to execute.
2328 5. There are several places where Exim runs a non-Exim command in a
2329 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2330 was being done only for the command run by the queryprogram router. It is
2331 now done for all such subprocesses. The other cases are: ${run, transport
2332 filters, and the commands run by the lmtp and pipe transports.
2334 6. Added CONFIGURE_GROUP build-time option.
2336 7. Some older OS have a limit of 256 on the maximum number of file
2337 descriptors. Exim was using setrlimit() to set 1000 as a large value
2338 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2339 systems. I've change it so that if it can't get 1000, it tries for 256.
2341 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2342 was an oversight, and furthermore, ever since the addition of extra
2343 controls (e.g. 4.43/32), the checks on when to allow different forms of
2344 "control" were broken. There should now be diagnostics for all cases when a
2345 control that does not make sense is encountered.
2347 9. Added the /retain_sender option to "control=submission".
2349 10. $recipients is now available in the predata ACL (oversight).
2351 11. Tidy the search cache before the fork to do a delivery from a message
2352 received from the command line. Otherwise the child will trigger a lookup
2353 failure and thereby defer the delivery if it tries to use (for example) a
2354 cached ldap connection that the parent has called unbind on.
2356 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2357 of $address_data from the recipient verification was clobbered by the
2358 sender verification.
2360 13. The value of address_data from a sender verification is now available in
2361 $sender_address_data in subsequent conditions in the ACL statement.
2363 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2365 15. Added a new option "connect=<time>" to callout options, to set a different
2368 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2369 was its contents. (It was OK if the option was not defined at all.)
2371 17. A "Completed" log line is now written for messages that are removed from
2372 the spool by the -Mrm option.
2374 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2375 information about exactly what failed.
2377 19. Added -dd to debug only the daemon process.
2379 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2380 handles timeouts, both on the server side and network timeouts. Renamed the
2381 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2383 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2384 It is now set to "smtps".
2386 22. $host_address is now set to the target address during the checking of
2387 ignore_target_hosts.
2389 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2390 being passed; this would have caused $sender_host_name to have been used if
2391 matching the list had actually called for a host name (not very likely,
2392 since this list is usually IP addresses). A host name is now passed as
2395 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2396 code that specifies a non-restarting handler (typically sigaction() in
2397 modern systems) in an attempt to fix a rare and obscure crash bug.
2399 25. Narrowed the window for a race in the daemon that could cause it to ignore
2400 SIGCHLD signals. This is not a major problem, because they are used only to
2401 wake it up if nothing else does.
2403 26. A malformed maildirsize file could cause Exim to calculate negative values
2404 for the mailbox size or file count. Odd effects could occur as a result.
2405 The maildirsize information is now recalculated if the size or filecount
2408 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2409 support for a long time. Removed HAVE_SYS_VFS_H.
2411 28. Installed the latest version of exipick from John Jetmore.
2413 29. In an address list, if the pattern was not a regular expression, an empty
2414 subject address (from a bounce message) matched only if the pattern was an
2415 empty string. Non-empty patterns were not even tested. This was the wrong
2416 because it is perfectly reasonable to use an empty address as part of a
2417 database query. An empty address is now tested by patterns that are
2418 lookups. However, all the other forms of pattern expect the subject to
2419 contain a local part and a domain, and therefore, for them, an empty
2420 address still always fails if the pattern is not itself empty.
2422 30. Exim went into a mad DNS loop when attempting to do a callout where the
2423 host was specified on an smtp transport, and looking it up yielded more
2424 than one IP address.
2426 31. Re-factored the code for checking spool and log partition space into a
2427 function that finds that data and another that does the check. The former
2428 is then used to implement four new variables: $spool_space, $log_space,
2429 $spool_inodes, and $log_inodes.
2431 32. The RFC2047 encoding function was originally intended for short strings
2432 such as real names; it was not keeping to the 75-character limit for
2433 encoded words that the RFC imposes. It now respects the limit, and
2434 generates multiple encoded words if necessary. To be on the safe side, I
2435 have increased the buffer size for the ${rfc2047: expansion operator from
2438 33. It is now permitted to omit both strings after an "if" condition; if the
2439 condition is true, the result is "true". As before, when the second string
2440 is omitted, a false condition yields an empty string. This makes it less
2441 cumbersome to write custom ACL and router conditions.
2443 34. Failure to deliver a bounce message always caused it to be frozen, even if
2444 there was an errors_to setting on the router. The errors_to setting is now
2447 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2448 canonical form (fully expanded) before being placed in
2449 $sender_host_address.
2451 36. The table in the code that translates DNS record types into text (T_A to
2452 "A" for instance) was missing entries for NS and CNAME. It is just possible
2453 that this could have caused confusion if both these types were looked up
2454 for the same domain, because the text type is used as part of Exim's
2455 per-process caching. But the chance of anyone hitting this buglet seems
2458 37. The dnsdb lookup has been extended in a number of ways.
2460 (1) There is a new type, "zns", which walks up the domain tree until it
2461 finds some nameserver records. It should be used with care.
2463 (2) There is a new type, "mxh", which is like "mx" except that it returns
2464 just the host names, not the priorities.
2466 (3) It is now possible to give a list of domains (or IP addresses) to be
2467 looked up. The behaviour when one of the lookups defers can be
2468 controlled by a keyword.
2470 (4) It is now possible to specify the separator character for use when
2471 multiple records are returned.
2473 38. The dnslists ACL condition has been extended: it is now possible to supply
2474 a list of IP addresses and/or domains to be looked up in a particular DNS
2477 39. Added log_selector=+queue_time_overall.
2479 40. When running the queue in the test harness, wait just a tad after forking a
2480 delivery process, to get repeatability of debugging output.
2482 41. Include certificate and key file names in error message when GnuTLS fails
2483 to set them up, because the GnuTLS error message doesn't include the name
2484 of the failing file when there is a problem reading it.
2486 42. Allow both -bf and -bF in the same test run.
2488 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2490 44. The "Exiscan patch" is now merged into the mainline Exim source.
2492 45. Sometimes the final signoff response after QUIT could fail to get
2493 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2494 < 0 before doing a fflush(). This bug looks as though it goes back to the
2495 introduction of TLS in release 3.20, but "sometimes" must have been rare
2496 because the tests only now provoked it.
2498 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2499 (this can affect the format of dates).
2501 47. exim_tidydb, when checking for the continued existence of a message for
2502 which it has found a message-specific retry record, was not finding
2503 messages that were in split spool directories. Consequently, it was
2504 deleting retry records that should have stayed in existence.
2506 48. Steve fixed some bugs in eximstats.
2508 49. The SPA authentication driver was not abandoning authentication and moving
2509 on to the next authenticator when an expansion was forced to fail,
2510 contradicting the general specification for all authenticators. Instead it
2511 was generating a temporary error. It now behaves as specified.
2513 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2514 (the order specifies the preference for clients). The order is now AES256,
2515 AES128, 3DES, ARCFOUR128.
2517 51. Small patch to Sieve code - explicitly set From: when generating an
2520 52. Exim crashed if a remote delivery caused a very long error message to be
2521 recorded - for instance if somebody sent an entire SpamAssassin report back
2522 as a large number of 550 error lines. This bug was coincidentally fixed by
2523 increasing the size of one of Exim's internal buffers (big_buffer) that
2524 happened as part of the Exiscan merge. However, to be on the safe side, I
2525 have made the code more robust (and fixed the comments that describe what
2528 53. Now that there can be additional text after "Completed" in log lines (if
2529 the queue_time_overall log selector is set), a one-byte patch to exigrep
2530 was needed to allow it to recognize "Completed" as not the last thing in
2533 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2534 patch that reportedly fixes this has been added. I am not expert enough to
2535 create a test for it. This is what the patch creator wrote:
2537 "I found a little strange behaviour of ldap code when working with
2538 Windows 2003 AD Domain, where users was placed in more than one
2539 Organization Units. When I tried to give exim partial DN, the exit code
2540 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2541 But simultaneously result of request was absolutely normal ldap result,
2542 so I produce this patch..."
2544 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2545 REFERENCE, so I have modified the code to exclude the patch when that macro
2548 55. Some experimental protocols are using DNS PTR records for new purposes. The
2549 keys for these records are domain names, not reversed IP addresses. The
2550 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2551 leaves it alone. Component reversal etc. now happens only for IP addresses.
2554 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2556 57. Double the size of the debug message buffer (to 2048) so that more of very
2557 long debug lines gets shown.
2559 58. The exicyclog utility now does better if the number of log files to keep
2560 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2562 59. Two changes related to the smtp_active_hostname option:
2564 (1) $smtp_active_hostname is now available as a variable.
2565 (2) The default for smtp_banner uses $smtp_active_hostname instead
2566 of $primary_hostname.
2568 60. The host_aton() function is supposed to be passed a string that is known
2569 to be a valid IP address. However, in the case of IPv6 addresses, it was
2570 not checking this. This is a hostage to fortune. Exim now panics and dies
2571 if the condition is not met. A case was found where this could be provoked
2572 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2573 components; fortuitously, this particular loophole had already been fixed
2574 by change 4.50/55 above.
2576 If there are any other similar loopholes, the new check in host_aton()
2577 itself should stop them being exploited. The report I received stated that
2578 data on the command line could provoke the exploit when Exim was running as
2579 exim, but did not say which command line option was involved. All I could
2580 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2581 running as the user.
2584 61. There was a buffer overflow vulnerability in the SPA authentication code
2585 (which came originally from the Samba project). I have added a test to the
2586 spa_base64_to_bits() function which I hope fixes it.
2589 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2590 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2592 63. The daemon start-up calls getloadavg() while still root for those OS that
2593 need the first call to be done as root, but it missed one case: when
2594 deliver_queue_load_max is set with deliver_drop_privilege. This is
2595 necessary for the benefit of the queue runner, because there is no re-exec
2596 when deliver_drop_privilege is set.
2598 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2599 This has been fixed.
2601 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2602 was in use, was not putting the data itself into the right store pool;
2603 consequently, it could be overwritten for a subsequent message in the same
2604 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2607 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2609 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2610 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2611 the function were treating the return as a boolean value, which happened to
2612 work because 0=false and not-0=true, but is not correct code.
2614 68. The host_aton() function was not handling scoped IPv6 addresses (those
2615 with, for example, "%eth0" on the end) correctly.
2617 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2618 negated items (that is, ~something) in unsigned ints. Some compilers
2619 apparently mutter when there is no cast.
2621 70. If an address verification called from an ACL failed, and did not produce a
2622 user-specific message (i.e. there was only a "system" message), nothing was
2623 put in $acl_verify_message. In this situation, it now puts the system
2626 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2627 session; change 4.31/43 added the unwanted input to the log line - except
2628 that it did not do this in the start of session case. It now does.
2630 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2631 This is wrong and can cause the other end to generate a synchronization
2632 error if it is another Exim or anything else that does the synchronization
2633 check. A QUIT command is no longer sent after a timeout.
2635 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2636 during host lookups.
2638 74. The defer_ok option of callout verification was not working if it was used
2639 when verifying addresses in header lines, that is, for this case:
2641 verify = header_sender/callout=defer_ok
2643 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2644 those file descriptors could be used for SMTP connections. If anything
2645 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2646 could be sent to the SMTP client, causing chaos. The daemon now opens
2647 stdin, stdout, and stderr to /dev/null when it puts itself into the
2650 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2651 log by default. The user can override this with suitable Perl magic.
2653 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2654 the log message when discard triggers, was not working for the DATA ACL or
2655 for the non-SMTP ACL.
2657 78. Error message wording change in sieve.c.
2659 79. If smtp_accept_max_per_host was set, the number of connections could be
2660 restricted to fewer than expected, because the daemon was trying to set up
2661 a new connection before checking whether the processes handling previous
2662 connections had finished. The check for completed processes is now done
2663 earlier. On busy systems, this bug wouldn't be noticed because something
2664 else would have woken the daemon, and it would have reaped the completed
2667 80. If a message was submitted locally by a user whose login name contained one
2668 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2669 It caused a spool format error. I have fixed the spool reading code. A
2670 related problem was that the "from" clause in the Received: line became
2671 illegal because of the space(s). It is now covered by ${quote_local_part.
2673 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2676 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2678 83. Patch from Sieve maintainer to fix unterminated string problem in
2679 "vacation" handling.
2681 84. Some minor changes to the Linux configuration files to help with other
2682 OS variants using glibc.
2684 85. One more patch for Sieve to update vacation handling to latest spec.
2687 ----------------------------------------------------
2688 See the note above about the 4.44 and 4.50 releases.
2689 ----------------------------------------------------
2695 1. Change 4.43/35 introduced a bug that caused file counts to be
2696 incorrectly computed when quota_filecount was set in an appendfile
2699 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2700 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2703 3. Give more explanation in the error message when the command for a transport
2704 filter fails to execute.
2706 4. There are several places where Exim runs a non-Exim command in a
2707 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2708 was being done only for the command run by the queryprogram router. It is
2709 now done for all such subprocesses. The other cases are: ${run, transport
2710 filters, and the commands run by the lmtp and pipe transports.
2712 5. Some older OS have a limit of 256 on the maximum number of file
2713 descriptors. Exim was using setrlimit() to set 1000 as a large value
2714 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2715 systems. I've change it so that if it can't get 1000, it tries for 256.
2717 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2718 was an oversight, and furthermore, ever since the addition of extra
2719 controls (e.g. 4.43/32), the checks on when to allow different forms of
2720 "control" were broken. There should now be diagnostics for all cases when a
2721 control that does not make sense is encountered.
2723 7. $recipients is now available in the predata ACL (oversight).
2725 8. Tidy the search cache before the fork to do a delivery from a message
2726 received from the command line. Otherwise the child will trigger a lookup
2727 failure and thereby defer the delivery if it tries to use (for example) a
2728 cached ldap connection that the parent has called unbind on.
2730 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2731 of $address_data from the recipient verification was clobbered by the
2732 sender verification.
2734 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2735 was its contents. (It was OK if the option was not defined at all.)
2737 11. A "Completed" log line is now written for messages that are removed from
2738 the spool by the -Mrm option.
2740 12. $host_address is now set to the target address during the checking of
2741 ignore_target_hosts.
2743 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2744 being passed; this would have caused $sender_host_name to have been used if
2745 matching the list had actually called for a host name (not very likely,
2746 since this list is usually IP addresses). A host name is now passed as
2749 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2750 code that specifies a non-restarting handler (typically sigaction() in
2751 modern systems) in an attempt to fix a rare and obscure crash bug.
2753 15. Narrowed the window for a race in the daemon that could cause it to ignore
2754 SIGCHLD signals. This is not a major problem, because they are used only to
2755 wake it up if nothing else does.
2757 16. A malformed maildirsize file could cause Exim to calculate negative values
2758 for the mailbox size or file count. Odd effects could occur as a result.
2759 The maildirsize information is now recalculated if the size or filecount
2762 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2763 support for a long time. Removed HAVE_SYS_VFS_H.
2765 18. Updated exipick to current release from John Jetmore.
2767 19. Allow an empty sender to be matched against a lookup in an address list.
2768 Previously the only cases considered were a regular expression, or an
2771 20. Exim went into a mad DNS lookup loop when doing a callout where the
2772 host was specified on the transport, if the DNS lookup yielded more than
2775 21. The RFC2047 encoding function was originally intended for short strings
2776 such as real names; it was not keeping to the 75-character limit for
2777 encoded words that the RFC imposes. It now respects the limit, and
2778 generates multiple encoded words if necessary. To be on the safe side, I
2779 have increased the buffer size for the ${rfc2047: expansion operator from
2782 22. Failure to deliver a bounce message always caused it to be frozen, even if
2783 there was an errors_to setting on the router. The errors_to setting is now
2786 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2787 canonical form (fully expanded) before being placed in
2788 $sender_host_address.
2790 24. Updated eximstats to version 1.33
2792 25. Include certificate and key file names in error message when GnuTLS fails
2793 to set them up, because the GnuTLS error message doesn't include the name
2794 of the failing file when there is a problem reading it.
2796 26. Expand error message when OpenSSL has problems setting up cert/key files.
2799 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2800 (this can affect the format of dates).
2802 28. exim_tidydb, when checking for the continued existence of a message for
2803 which it has found a message-specific retry record, was not finding
2804 messages that were in split spool directories. Consequently, it was
2805 deleting retry records that should have stayed in existence.
2807 29. eximstats updated to version 1.35
2808 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2809 1.35 - bugfix such that pie charts by volume are generated correctly
2811 30. The SPA authentication driver was not abandoning authentication and moving
2812 on to the next authenticator when an expansion was forced to fail,
2813 contradicting the general specification for all authenticators. Instead it
2814 was generating a temporary error. It now behaves as specified.
2816 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2817 (the order specifies the preference for clients). The order is now AES256,
2818 AES128, 3DES, ARCFOUR128.
2820 31. Small patch to Sieve code - explicitly set From: when generating an
2823 32. Exim crashed if a remote delivery caused a very long error message to be
2824 recorded - for instance if somebody sent an entire SpamAssassin report back
2825 as a large number of 550 error lines. This bug was coincidentally fixed by
2826 increasing the size of one of Exim's internal buffers (big_buffer) that
2827 happened as part of the Exiscan merge. However, to be on the safe side, I
2828 have made the code more robust (and fixed the comments that describe what
2831 33. Some experimental protocols are using DNS PTR records for new purposes. The
2832 keys for these records are domain names, not reversed IP addresses. The
2833 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2834 leaves it alone. Component reversal etc. now happens only for IP addresses.
2837 34. The host_aton() function is supposed to be passed a string that is known
2838 to be a valid IP address. However, in the case of IPv6 addresses, it was
2839 not checking this. This is a hostage to fortune. Exim now panics and dies
2840 if the condition is not met. A case was found where this could be provoked
2841 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2842 components; fortuitously, this particular loophole had already been fixed
2843 by change 4.50/55 or 4.44/33 above.
2845 If there are any other similar loopholes, the new check in host_aton()
2846 itself should stop them being exploited. The report I received stated that
2847 data on the command line could provoke the exploit when Exim was running as
2848 exim, but did not say which command line option was involved. All I could
2849 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2850 running as the user.
2853 35. There was a buffer overflow vulnerability in the SPA authentication code
2854 (which came originally from the Samba project). I have added a test to the
2855 spa_base64_to_bits() function which I hope fixes it.
2858 36. The daemon start-up calls getloadavg() while still root for those OS that
2859 need the first call to be done as root, but it missed one case: when
2860 deliver_queue_load_max is set with deliver_drop_privilege. This is
2861 necessary for the benefit of the queue runner, because there is no re-exec
2862 when deliver_drop_privilege is set.
2864 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2865 was in use, was not putting the data itself into the right store pool;
2866 consequently, it could be overwritten for a subsequent message in the same
2867 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2870 38. Sometimes the final signoff response after QUIT could fail to get
2871 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2872 < 0 before doing a fflush(). This bug looks as though it goes back to the
2873 introduction of TLS in release 3.20, but "sometimes" must have been rare
2874 because the tests only now provoked it.
2880 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2881 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2882 Now it returns an integer. A number of calls were still expecting a T/F
2883 return. Fortuitously, in all cases, the tests worked in OK situations,
2884 which is the norm. However, things would have gone wrong on any write
2885 failures on the smtp file descriptor. This function is used when sending
2886 messages over SMTP and also when doing verify callouts.
2888 2. When Exim is called to do synchronous delivery of a locally submitted
2889 message (the -odf or -odi options), it no longer closes stderr before doing
2892 3. Implemented the mua_wrapper option.
2894 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2896 5. Implemented the functions header_remove(), header_testname(),
2897 header_add_at_position(), and receive_remove_recipient(), and exported them
2900 6. If an ACL "warn" statement specified the addition of headers, Exim already
2901 inserted X-ACL-Warn: at the start if there was no header name. However, it
2902 was not making this test for the second and subsequent header lines if
2903 there were newlines in the string. This meant that an invalid header could
2904 be inserted if Exim was badly configured.
2906 7. Allow an ACL "warn" statement to add header lines at the start or after all
2907 the Received: headers, as well as at the end.
2909 8. Added the rcpt_4xx retry error code.
2911 9. Added postmaster_mailfrom=xxx to callout verification option.
2913 10. Added mailfrom=xxxx to the callout verification option, for verify=
2916 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2917 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2919 12. Inserted some casts to stop certain compilers warning when using pointer
2920 differences as field lengths or precisions in printf-type calls (mostly
2921 affecting debugging statements).
2923 13. Added optional readline() support for -be (dynamically loaded).
2925 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2926 same clock tick as a message's arrival, so that its received time was the
2927 same as the "first fail" time on the retry record, and that message
2928 remained on the queue past the ultimate address timeout, every queue runner
2929 would try a delivery (because it was past the ultimate address timeout) but
2930 after another failure, the ultimate address timeout, which should have then
2931 bounced the address, did not kick in. This was a "< instead of <=" error;
2932 in most cases the first failure would have been in the next clock tick
2933 after the received time, and all would be well.
2935 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2936 being recognized when the domain list was tested by the match_domain
2937 condition in an expansion string.
2939 16. Added the ${str2b64: operator.
2941 17. Exim was always calling setrlimit() to set a large limit for the number of
2942 processes, without checking whether the existing limit was already
2943 adequate. (It did check for the limit on file descriptors.) Furthermore,
2944 errors from getrlimit() and setrlimit() were being ignored. Now they are
2945 logged to the main and panic logs, but Exim does carry on, to try to do its
2946 job under whatever limits there are.
2948 18. Imported PCRE 5.0.
2950 19. Trivial typo in log message " temporarily refused connection" (the leading
2953 20. If the log selector return_path_on_delivery was set and an address was
2954 redirected to /dev/null, the delivery process crashed because it assumed
2955 that a return path would always be set for a "successful" delivery. In this
2956 case, the whole delivery is bypassed as an optimization, and therefore no
2959 21. Internal re-arrangement: the function for sending a challenge and reading
2960 a response while authentication was assuming a zero-terminated challenge
2961 string. It's now changed to take a pointer and a length, to allow for
2962 binary data in such strings.
2964 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2966 23. Exim was not respecting finduser_retries when seeking the login of the
2967 uid under which it was called; it was always trying 10 times. (The default
2968 setting of finduser_retries is zero.) Also, it was sleeping after the final
2969 failure, which is pointless.
2971 24. Implemented tls_on_connect_ports.
2973 25. Implemented acl_smtp_predata.
2975 26. If the domain in control=submission is set empty, Exim assumes that the
2976 authenticated id is a complete email address when it generates From: or
2977 Sender: header lines.
2979 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2980 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2981 chown and chgrp in /bin and hostname in /usr/bin.
2983 28. Exim was keeping the "process log" file open after each use, just as it
2984 does for the main log. This opens the possibility of it remaining open for
2985 long periods when the USR1 signal hits a daemon. Occasional processlog
2986 errors were reported, that could have been caused by this. Anyway, it seems
2987 much more sensible not to leave this file open at all, so that is what now
2990 29. The long-running daemon process does not normally write to the log once it
2991 has entered its main loop, and it closes the log before doing so. This is
2992 so that log files can straightforwardly be renamed and moved. However,
2993 there are a couple of unusual error situations where the daemon does write
2994 log entries, and I had neglected to close the log afterwards.
2996 30. The text of an SMTP error response that was received during a remote
2997 delivery was being truncated at 512 bytes. This is too short for some of
2998 the long messages that one sometimes sees. I've increased the limit to
3001 31. It is now possible to make retry rules that apply only when a message has a
3002 specific sender, in particular, an empty sender.
3004 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3005 it possible to be selective about when SMTP synchronization is enforced.
3007 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3009 32. Implemented hosts_connection_nolog.
3011 33. Added an ACL for QUIT.
3013 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3016 35. Added mailbox_size and mailbox_filecount to appendfile.
3018 36. Added control = no_multiline_responses to ACLs.
3020 37. There was a bug in the logic of the code that waits for the clock to tick
3021 in the case where the clock went backwards by a substantial amount such
3022 that the microsecond fraction of "now" was more than the microsecond
3023 fraction of "then" (but the whole seconds number was less).
3025 38. Added support for the libradius Radius client library this is found on
3026 FreeBSD (previously only the radiusclient library was supported).
3032 1. When certain lookups returned multiple values in the form name=value, the
3033 quoting of the values was not always being done properly. Specifically:
3034 (a) If the value started with a double quote, but contained no whitespace,
3036 (b) If the value contained whitespace other than a space character (i.e.
3037 tabs or newlines or carriage returns) it was not quoted.
3038 This fix has been applied to the mysql and pgsql lookups by writing a
3039 separate quoting function and calling it from the lookup code. The fix
3040 should probably also be applied to nisplus, ibase and oracle lookups, but
3041 since I cannot test any of those, I have not disturbed their existing code.
3043 2. A hit in the callout cache for a specific address caused a log line with no
3044 reason for rejecting RCPT. Now it says "Previous (cached) callout
3045 verification failure".
3047 3. There was an off-by-one bug in the queryprogram router. An over-long
3048 return line was truncated at 256 instead of 255 characters, thereby
3049 overflowing its buffer with the terminating zero. As well as fixing this, I
3050 have increased the buffer size to 1024 (and made a note to document this).
3052 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3053 when Exim is waiting for an SMTP response from a remote server, Exim
3054 restarts its select() call on the socket, thereby resetting its timeout.
3055 This is not a problem when such interrupts are rare. Somebody set up a cron
3056 job to run exiwhat every 2 minutes, which is less than the normal select()
3057 timeout (5 or 10 minutes). This meant that the select() timeout never
3058 kicked in because it was always reset. I have fixed this by comparing the
3059 time when an interrupt arrives with the time at the start of the first call
3060 to select(). If more time than the timeout has elapsed, the interrupt is
3061 treated as a timeout.
3063 5. Some internal re-factoring in preparation for the addition of Sieve
3064 extensions (by MH). In particular, the "personal" test is moved to a
3065 separate function, and given an option for scanning Cc: and Bcc: (which is
3066 not set for Exim filters).
3068 6. When Exim created an email address using the login of the caller as the
3069 local part (e.g. when creating a From: or Sender: header line), it was not
3070 quoting the local part when it contained special characters such as @.
3072 7. Installed new OpenBSD configuration files.
3074 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3075 try to make them clearer.
3077 9. Callout options, other than the timeout value, were being ignored when
3078 verifying sender addresses in header lines. For example, when using
3080 verify = header_sender/callout=no_cache
3082 the cache was (incorrectly) being used.
3084 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3085 only the Cygwin environment.
3087 11. When return_path_on_delivery was set as a log selector, if different remote
3088 addresses in the same message used different return paths and parallel
3089 remote delivery occurred, the wrong values would sometimes be logged.
3090 (Whenever a remote delivery process finished, the return path value from
3091 the most recently started remote delivery process was logged.)
3093 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3094 lines when AUTH and/or TLS are in use. This is the "received protocol"
3095 field. Exim used to use "asmtp" for authenticated SMTP, without any
3096 indication (in the protocol name) for TLS use. Now it follows the RFC and
3097 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3098 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3099 names appear in log lines as well as in Received: header lines.
3101 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3102 extensions, and comparison tests, and to fix some bugs.
3104 14. Changes to the "personal" filter test:
3106 (1) The test was buggy in that it was just doing the equivalent of
3107 "contains" tests on header lines. For example, if a user's address was
3108 anne@some.where, the "personal" test would incorrectly be true for
3110 To: susanne@some.where
3112 This test is now done by extracting each address from the header in turn,
3113 and checking the entire address. Other tests that are part of "personal"
3114 are now done using regular expressions (for example, to check local parts
3115 of addresses in From: header lines).
3117 (2) The list of non-personal local parts in From: addresses has been
3118 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3119 taken from the Sieve specification recommendations.
3121 (3) If the message contains any header line starting with "List-" it is
3122 treated as non-personal.
3124 (4) The test for "circular" in the Subject: header line has been removed
3125 because it now seems ill-conceived.
3127 15. Minor typos in src/EDITME comments corrected.
3129 16. Installed latest exipick from John Jetmore.
3131 17. If headers_add on a router specified a text string that was too long for
3132 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3133 of string_sprintf() is now avoided.
3135 18. $message_body_size was not set (it was always zero) when running the DATA
3136 ACL and the local_scan() function.
3138 19. For the "mail" command in an Exim filter, no default was being set for
3139 the once_repeat time, causing a random time value to be used if "once" was
3140 specified. (If the value happened to be <= 0, no repeat happened.) The
3141 default is now 0s, meaning "never repeat". The "vacation" command was OK
3142 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3143 (I found it when inspecting the code).
3145 20. There is now an overall timeout for performing a callout verification. It
3146 defaults to 4 times the callout timeout, which applies to individual SMTP
3147 commands during the callout. The overall timeout applies when there is more
3148 than one host that can be tried. The timeout is checked before trying the
3149 next host. This prevents very long delays if there are a large number of
3150 hosts and all are timing out (e.g. when the network connections are timing
3151 out). The value of the overall timeout can be changed by specifying an
3152 additional sub-option for "callout", called "maxwait". For example:
3154 verify = sender/callout=5s,maxwait=20s
3156 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3157 to the end before writing, but this should make it even safer).
3159 22. Exim was forgetting that it had advertised PIPELINING for the second and
3160 subsequent messages on an SMTP connection. It was also not resetting its
3161 memory on STARTTLS and an internal HELO.
3163 23. When Exim logs an SMTP synchronization error within a session, it now
3164 records whether PIPELINING has been advertised or not.
3166 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3167 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3168 rather than long int.
3170 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3172 26. Added the never_mail option to autoreply.
3178 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3179 crash if the getsockname() call failed; this can happen if a connection is
3180 closed very soon after it is established. The problem was simply in the
3181 order in which certain operations were done, causing Exim to try to write
3182 to the SMTP stream before it had set up the file descriptor. The bug has
3183 been fixed by making things happen in the correct order.
3189 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3190 before the connection was closed, thus losing the rejection response.
3192 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3193 some early Solaris releases, but causes trouble in current releases where
3194 socklen_t is defined.
3196 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3199 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3202 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3203 that releasing the top of it at the end releases what was used for sub-
3204 expansions (unless the block got too big). However, discard this block if
3205 the first thing is a variable or header, so that we can use its block when
3206 it is dynamic (useful for very large $message_headers, for example).
3208 6. Lookups now cache *every* query, not just the most recent. A new, separate
3209 store pool is used for this. It can be recovered when all lookup caches are
3210 flushed. Lookups now release memory at the end of their result strings.
3211 This has involved some general refactoring of the lookup sources.
3213 7. Some code has been added to the store_xxx() functions to reduce the amount
3214 of flapping under certain conditions.
3216 8. log_incoming_interface used to affect only the <= reception log lines. Now
3217 it causes the local interface and port to be added to several more SMTP log
3218 lines, for example "SMTP connection from", and rejection lines.
3220 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3222 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3224 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3225 could be overwritten at the end of the current message (or the start of a
3226 new message if it was set in a HELO ACL). The value is now preserved for
3227 the duration of the SMTP connection.
3229 12. If a transport had a headers_rewrite setting, and a matching header line
3230 contained an unqualified address, that address was qualified, even if it
3231 did not match any rewriting rules. The underlying bug was that the values
3232 of the flags that permit the existence of unqualified sender and recipient
3233 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3234 non-local messages, and by -bnq for local messages) were not being
3235 preserved with the message after it was received.
3237 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3238 "next input=" as part of the text comprising the host identity instead of
3239 the correct text. The code was using the same buffer for two different
3240 strings. However, depending on which order the printing function evaluated
3241 its arguments, the bug did not always show up. Under Linux, for example, my
3242 test suite worked just fine.
3244 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3245 doesn't work with some older versions of Perl. It has been changed to "my",
3246 which in any case is probably the better facility to use.
3248 15. A really picky compiler found some instances of statements for creating
3249 error messages that either had too many or two few arguments for the format
3252 16. The size of the buffer for calls to the DNS resolver has been increased
3253 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3254 for addresses that have a lot of PTR records. This alleviates a problem; it
3255 does not fully solve it.
3257 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3258 buffer now truncates the list and logs the incident, which is the same
3259 action as happens when Exim is looking up a host name and its aliases.
3260 Previously in this situation something unpredictable would happen;
3261 sometimes it was "internal error: store_reset failed".
3263 18. If a server dropped the connection unexpectedly when an Exim client was
3264 using GnuTLS and trying to read a response, the client delivery process
3265 crashed while trying to generate an error log message.
3267 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3268 string, for example:
3270 warn message = H1: something\nH2: something
3272 the text was added as a single header line from Exim's point of view
3273 though it ended up OK in the delivered message. However, searching for the
3274 second and subsequent header lines using $h_h2: did not work. This has been
3275 fixed. Similarly, if a system filter added multiple headers in this way,
3276 the routers could not see them.
3278 20. Expanded the error message when iplsearch is called with an invalid key to
3279 suggest using net-iplsearch in a host list.
3281 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3282 ACLs are no longer actually imposed (and a message to that effect is
3285 22. If a "gecos" field in a passwd entry contained escaped characters, in
3286 particular, if it contained a \" sequence, Exim got it wrong when building
3287 a From: or a Sender: header from that name. A second bug also caused
3288 incorrect handling when an unquoted " was present following a character
3289 that needed quoting.
3291 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3292 was not being matched caselessly.
3294 24. Arranged for all hyphens in the exim.8 source to be escaped with
3297 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3298 properly. Recipient callout cache records were still being keyed to include
3299 the sender, even when use_sender was set false. This led to far more
3300 callouts that were necessary. The sender is no longer included in the key
3301 when use_sender is false.
3303 26. Added "control = submission" modifier to ACLs.
3305 27. Added the ${base62d: operator to decode base 62 numbers.
3307 28. dnsdb lookups can now access SRV records.
3309 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3310 the configuration file.
3312 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3313 (-v) mode. This makes the output for a verbose queue run more intelligible.
3315 31. Added a use_postmaster feature to recipient callouts.
3317 32. Added the $body_zerocount variable, containing the number of binary zero
3318 bytes in the message body.
3320 33. The time of last modification of the "new" subdirectory is now used as the
3321 "mailbox time last read" when there is a quota error for a maildir
3324 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3326 35. Added +ignore_unknown as a special item in host lists.
3328 36. Code for decoding IPv6 addresses in host lists is now included, even if
3329 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3330 address was recognized as an IP address, but was then not correctly decoded
3331 into binary, causing unexpected and incorrect effects when compared with
3338 1. Very minor rewording of debugging text in manualroute to say "list of
3339 hosts" instead of "hostlist".
3341 2. If verify=header_syntax was set, and a header line with an unqualified
3342 address (no domain) and a large number of spaces between the end of the
3343 name and the colon was received, the reception process suffered a buffer
3344 overflow, and (when I tested it) crashed. This was caused by some obsolete
3345 code that should have been removed. The fix is to remove it!
3347 3. When running in the test harness, delay a bit after writing a bounce
3348 message to get a bit more predictability in the log output.
3350 4. Added a call to search_tidyup() just before forking a reception process. In
3351 theory, someone could use a lookup in the expansion of smtp_accept_max_
3352 per_host which, without the tidyup, could leave open a database connection.
3354 5. Added the variables $recipient_data and $sender_data which get set from a
3355 lookup success in an ACL "recipients" or "senders" condition, or a router
3356 "senders" option, similar to $domain_data and $local_part_data.
3358 6. Moved the writing of debug_print from before to after the "senders" test
3361 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3362 problems for message scanning, either using a data ACL, or using
3363 local_scan() because the Received: header was not generated till after they
3364 were called (in order to set the time as the time of reception completion).
3365 I have revised the way this works. The header is now generated after the
3366 body is received, but before the ACL or local_scan() are called. After they
3367 are run, the timestamp in the header is updated.
3373 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3374 before starting a queue runner without re-exec. This happened only when
3375 deliver_drop_privilege was set or when the Exim user was set to root. The
3376 effect of the bug was that timeouts during subsequent deliveries caused
3377 crashes instead of being properly handled. The handler is now left at its
3378 default (and expected) setting.
3380 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3381 message, again when deliver_drop_privilege is set or Exim is run as root.
3382 The bug described in (1) was not present in this case, but the tidying up
3383 of the other signals was missing. I have made the two cases consistent.
3385 3. The ignore_target_hosts setting on a manualroute router was being ignored
3386 for hosts that were looked up using the /MX notation.
3388 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3391 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3392 operated on the sender address. After changing the $sender_address to <>
3393 for the sender address verify, Exim was re-instated it as the original
3394 (before rewriting) address, but remembering that it had rewritten it, so it
3395 wasn't rewriting it again. This bug also had the effect of breaking the
3396 sender address verification caching when the sender address was rewritten.
3398 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3399 This has been changed so that if the ip literal address matches
3400 ignore_target_hosts, the router declines.
3402 7. Added expansion conditions match_domain, match_address, and match_local_
3403 part (NOT match_host).
3405 8. The placeholder for the Received: header didn't have a length field set.
3407 9. Added code to Exim itself and to exim_lock to test for a specific race
3408 condition that could lead to file corruption when using MBX delivery. The
3409 issue is with the lockfile that is created in /tmp. If this file is removed
3410 after a process has opened it but before that process has acquired a lock,
3411 there is the potential for a second process to recreate the file and also
3412 acquire a lock. This could lead to two Exim processes writing to the file
3413 at the same time. The added code performs the same test as UW imapd; it
3414 checks after acquiring the lock that its file descriptor still refers to
3415 the same named file.
3417 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3418 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3421 11. Added the smtp_active_hostname option. If used, this will typically be made
3422 to depend on the incoming interface address. Because $interface_address is
3423 not set up until the daemon has forked a reception process, error responses
3424 that can happen earlier (such as "too many connections") no longer contain
3427 12. If an expansion in a condition on a "warn" statement fails because a lookup
3428 defers, the "warn" statement is abandoned, and the next ACL statement is
3429 processed. Previously this caused the whole ACL to be aborted.
3431 13. Added the iplsearch lookup type.
3433 14. Added ident_timeout as a log selector.
3435 15. Added tls_certificate_verified as a log selector.
3437 16. Added a global option tls_require_ciphers (compare the smtp transport
3438 option of the same name). This controls incoming TLS connections.
3440 17. I finally figured out how to make tls_require_ciphers do a similar thing
3441 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3442 before starting the TLS session.
3444 18. Tabs are now shown as \t in -bP output.
3446 19. If the log selector return_path_on_delivery was set, Exim crashed when
3447 bouncing a message because it had too many Received: header lines.
3449 20. If two routers both had headers_remove settings, and the first one included
3450 a superfluous trailing colon, the final name in the first list and the
3451 first name in the second list were incorrectly joined into one item (with a
3452 colon in the middle).
3458 1. Added -C and -D options to the exinext utility, mainly to make it easier
3459 to include in the automated testing, but these could be helpful when
3460 multiple configurations are in use.
3462 2. The exinext utility was not formatting the output nicely when there was
3463 an alternate port involved in the retry record key, nor when there was a
3464 message id as well (for retries that were specific to a specific message
3465 and a specific host). It was also confused by IPv6 addresses, because of
3466 the additional colons they contain. I have fixed the IPv4 problem, and
3467 patched it up to do a reasonable job for IPv6.
3469 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3470 delivery, the log line now contains "pipelined" if PIPELINING was used.
3472 4. An SMTP transport process used to panic and die if the bind() call to set
3473 an explicit outgoing interface failed. This has been changed; it is now
3474 treated in the same way as a connect() failure.
3476 5. A reference to $sender_host_name in the part of a conditional expansion
3477 that was being skipped was still causing a DNS lookup. This no longer
3480 6. The def: expansion condition was not recognizing references to header lines
3481 that used bh_ and bheader_.
3483 7. Added the _cache feature to named lists.
3485 8. The code for checking quota_filecount in the appendfile transport was
3486 allowing one more file than it should have been.
3488 9. For compatibility with Sendmail, the command line option
3496 and sets the incoming protocol and host name (for trusted callers). The
3497 host name and its colon can be omitted when only the protocol is to be set.
3498 Note the Exim already has two private options, -pd and -ps, that refer to
3499 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3500 "s", but I don't think that's a major issue.
3502 10. A number of refactoring changes to the code, none of which should affect
3505 (a) The number of logging options was getting close to filling up the
3506 32-bit word that was used as a bit map. I have split them into two classes:
3507 those that are passed in the argument to log_write(), and those that are
3508 only ever tested independently outside of that function. These are now in
3509 separate 32-bit words, so there is plenty of room for expansion again.
3510 There is no change in the user interface or the logging behaviour.
3512 (b) When building, for example, log lines, the code previously used a
3513 macro that called string_cat() twice, in order to add two strings. This is
3514 not really sufficiently general. Furthermore, there was one instance where
3515 it was actually wrong because one of the argument was used twice, and in
3516 one call a function was used. (As it happened, calling the function twice
3517 did not affect the overall behaviour.) The macro has been replaced by a
3518 function that can join an arbitrary number of extra strings onto a growing
3521 (c) The code for expansion conditions now uses a table and a binary chop
3522 instead of a serial search (which was left over from when there were very
3523 few conditions). Also, it now recognizes conditions like "pam" even when
3524 the relevant support is not compiled in: a suitably worded error message is
3525 given if an attempt is made to use such a condition.
3527 11. Added ${time_interval:xxxxx}.
3529 12. A bug was causing one of the ddress fields not to be passed back correctly
3530 from remote delivery subprocesses. The field in question was not being
3531 subsequently used, so this caused to problems in practice.
3533 13. Added new log selectors queue_time and deliver_time.
3535 14. Might have fixed a bug in maildirsizefile handling that threw up
3536 "unexpected character" debug warnings, and recalculated the data
3537 unnecessarily. In any case, I expanded the warning message to give more
3540 15. Added the message "Restricted characters in address" to the statements in
3541 the default ACL that block characters like @ and % in local parts.
3543 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3544 Three changes have been made:
3546 (a) There was a serious bug; a negative response to MAIL caused the whole
3547 recipient domain to be cached as invalid, thereby blocking all messages
3548 to all local parts at the same domain, from all senders. This bug has
3549 been fixed. The domain is no longer cached after a negative response to
3550 MAIL if the sender used is not empty.
3552 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3555 (c) A new callout option, "use_sender" has been added for people who want
3556 the modified behaviour.
3562 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3565 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3566 indeed breaks things for older releases.
3568 3. Added additional logging to the case where there is a problem reading data
3569 from a filter that is running in a subprocess using a pipe, in order to
3570 try to track down a specific problem.
3572 4. Testing facility fudge: when running in the test harness and attempting
3573 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3574 getting "No route to host". Convert this to a timeout.
3576 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3579 6. Some OS don't have socklen_t but use size_t instead. This affects the
3580 fifth argument of getsockopt() amongst other things. This is now
3581 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3582 can be set for individual OS. I have set it for SunOS5, OSF1, and
3583 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3584 some earlier ones do not.
3586 7. Change 4.30/15 was not doing the test caselessly.
3588 8. The standard form for an IPv6 address literal was being rejected by address
3589 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3590 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3591 this, as well as the form without the "IPv6" on the front (but only when
3592 address literals are enabled, of course).
3594 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3596 10. Exim crashed if a message with an empty sender address specified by -f
3597 encountered a router with an errors_to setting. This could be provoked only
3598 by a command such as
3602 where an empty string was supplied; "<>" did not hit this bug.
3604 11. Installed PCRE release 4.5.
3606 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3607 remained set. It is now erased.
3609 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3610 times from message ids (which are base 36 rather than the normal 62).
3612 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3613 were being counted as actual protocol errors, and logged if the log
3614 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3615 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3616 and DATA following a set of rejected RCPTs do not count as protocol errors.
3617 In other words, Exim assumes they were pipelined, though this may not
3618 actually be the case. Of course, in all cases the client gets an
3619 appropriate error code.
3621 15. If a lookup fails in an ACL condition, a message about the failure may
3622 be available; it is used if testing the ACL cannot continue, because most
3623 such messages specify what the cause of the deferral is. However, some
3624 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3625 that caused an old message to be retained and used if a later statement
3626 caused a defer, replacing the real cause of the deferral.
3628 16. If an IP address had so many PTR records that the DNS lookup buffer
3629 was not large enough to hold them, Exim could crash while trying to process
3630 the truncated data. It now detects and logs this case.
3632 17. Further to 4.21/58, another change has been made: if (and only if) the
3633 first line of a message (the first header line) ends with CRLF, a bare LF
3634 in a subsequent header line has a space inserted after it, so as not to
3635 terminate the header.
3637 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3638 unnecessarily, used atoi() instead of strtol(), and didn't check the
3639 termination when getting file sizes from file names by regex.
3641 19. Completely re-implemented the support for maildirsize files, in the light
3642 of a number of problems with the previous contributed implementation
3643 (4.30/29). In particular:
3645 . If the quota is zero, the maildirsize file is maintained, but no quota is
3648 . If the maildir directory does not exist, it is created before any attempt
3649 to write a maildirsize file.
3651 . The quota value in the file is just a cache; if the quota is changed in
3652 the transport, the new value overrides.
3654 . A regular expression is available for excluding directories from the
3657 20. The autoreply transport checks the characters in options that define the
3658 message's headers; it allows continued headers, but it was checking with
3659 isspace() after an embedded newline instead of explicitly looking for a
3662 21. If all the "regular" hosts to which an address was routed had passed their
3663 expiry times, and had not reached their retry times, the address was
3664 bounced, even if fallback hosts were defined. Now Exim should go on to try
3667 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3668 equivalent code in the SMTP transport. Some hosts send humungous responses
3669 to HELO/EHLO, more than 1024 it seems.
3671 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3672 gives compiler warnings in some environments. I've now done it "properly",
3675 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3676 (because of problems with the built-in one) was declared to return uschar *
3677 instead of char *, causing compiler failure.
3679 25. Fixed a file descriptor leak when processing alias/forward files.
3681 26. Fixed a minor format string issue in dbfn.c.
3683 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3685 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3686 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3689 29. When a sender address is verified, it is cached, to save repeating the test
3690 when there is more than one recipient in a message. However, when the
3691 verification involves a callout, it is possible for different callout
3692 options to be set for different recipients. It is too complicated to keep
3693 track of this in the cache, so now Exim always runs a verification when a
3694 callout is required, relying on the callout cache for the optimization.
3695 The overhead is duplication of the address routing, but this should not be
3698 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3699 to be verified with callout=postmaster, and the main callout worked but the
3700 postmaster check failed, the verification correctly failed. However, if a
3701 subsequent RCPT command asked for sender verification *without* the
3702 postmaster check, incorrect caching caused this verification also to fail,
3705 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3706 it was not caching the DNS options (qualify_single, search_parents) that
3707 were used when the lookup failed. A subsequent lookup with different
3708 options therefore always gave the same answer, though there were cases
3709 where it should not have. (Example: a "domains = !$mx_any" option on a
3710 dnslookup router: the "domains" option is always processed without any
3711 widening, but the router might have qualify_single set.) Now Exim uses the
3712 cached value only when the same options are set.
3714 32. Added John Jetmore's "exipick" utility to the distribution.
3716 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3717 than a timeout (e.g. a certificate is required, and is not provided), an
3718 Exim server now closes the connection immediately. Previously it waited for
3719 the client to close - but if the client is SSL, it seems that they each
3720 wait for each other, leading to a delay before one of them times out.
3722 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3723 maintained 0.8.x compatibility because I don't think many are using it, and
3724 it is clearly obsolete.
3726 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3729 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3730 client certificate was expired. A simple patch fixes this, though I don't
3731 understand the full logic of why the verify callback is called multiple
3734 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3735 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3736 which causes problems with some clients (such as the Certicom SSL Plus
3737 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3738 disables the coutermeasure allowing Eudora to connect."
3740 38. Exim was not checking that a write() to a log file succeeded. This could
3741 lead to Bad Things if a log got too big, in particular if it hit a file
3742 size limit. Exim now panics and dies if it cannot write to a log file, just
3743 as it does if it cannot open a log file.
3745 39. Modified OS/Makefile-Linux so that it now contains
3747 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3749 The two -D definitions ensure that Exim is compiled with large file
3750 support, which makes it possible to handle log files that are bigger than
3753 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3754 instance) a domain was checked against a named list that involved a lookup,
3755 causing $domain_data to be set, then another domain was checked against the
3756 same list, then the first domain was re-checked, the value of $domain_data
3757 after the final check could be wrong. In particular, if the second check
3758 failed, it could be set empty. This bug probably also applied to
3761 41. The strip_trailing_dot option was not being applied to the address given
3762 with the -f command-line option.
3764 42. The code for reading a message's header from the spool was incrementing
3765 $received_count, but never initializing it. This meant that the value was
3766 incorrect (doubled) while delivering a message in the same process in which
3767 it was received. In the most common configuration of Exim, this never
3768 happens - a fresh exec is done - but it can happen when
3769 deliver_drop_privilege is set.
3771 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3772 it now includes up to 150 characters of the unexpected data in the log
3775 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3776 and building data strings. The size of both of these buffers was 10 000
3777 bytes - far larger than anybody would *ever* want, thought I. Needless to
3778 say, somebody hit the limit. I have increased the maximum line length to
3779 20 000 and the maximum data length of concatenated lines to 100 000. I have
3780 also fixed two bugs, because there was no checking on these buffers. Tsk,
3781 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3782 buffer is too small.
3784 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3785 lsearch lookups. Now it does.
3787 46. When parsing a route_list item in a manualroute router, a fixed-length
3788 buffer was used for the list of hosts. I made this 1024 bytes long,
3789 thinking that nobody would ever have a list of hosts that long. Wrong.
3790 Somebody had a whole pile of complicated expansion conditions, and the
3791 string was silently truncated, leading to an expansion error. It turns out
3792 that it is easier to change to an unlimited length (owing to other changes
3793 that have happened since this code was originally written) than to build
3794 structure for giving a limitation error. The length of the item that
3795 expands into the list of hosts is now unlimited.
3797 47. The lsearch lookup could not handle data where the length of text line was
3798 more than 4095 characters. Such lines were truncated, leading to shortened
3799 data being returned. It should now handle lines of any length.
3801 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3802 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3805 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3806 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3807 no longer happens in comment lines. A list of replacements is now placed
3808 at the head of all of the source files, except those whose only change is
3809 to replace PERL_COMMAND in the very first #! line.
3811 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3812 messages on the queue, with a bottom-up merge sort, using code contributed
3813 by Michael Haardt. This should make operations like -bp somewhat faster on
3814 large queues. It won't affect queue runners, except when queue_run_in_order
3817 51. Installed eximstats 1.31 in the distribution.
3819 52. Added support for SRV lookups to the dnslookup router.
3821 53. If an ACL referred to $message_body or $message_body_end, the value was not
3822 reset for any messages that followed in the same SMTP session.
3824 54. The store-handling optimization for building very long strings was not
3825 differentiating between the different store pools. I don't think this
3826 actually made any difference in practice, but I've tidied it.
3828 55. While running the routers to verify a sender address, $sender_address
3829 was still set to the sender address. This is wrong, because when routing to
3830 send a bounce to the sender, it would be empty. Therefore, I have changed
3831 it so that, while verifying a sender address, $sender_address is set to <>.
3832 (There is no change to what happens when verifying a recipient address.)
3834 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3835 target A or AAAA records (if not already returned) without resetting the
3836 qualify_single or search_parents options of the DNS resolver. These are
3837 inappropriate in this case because the targets of MX and SRV records must
3838 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3839 target that, when qualified, matched something in the local domain. These
3840 two options are now turned off when doing these lookups.
3842 57. It seems that at least some releases of Reiserfs (which does not have the
3843 concept of a fixed number of inodes) returns zero and not -1 for the
3844 number of available inodes. This interacted badly with check_spool_inodes,
3845 which assumed that -1 was the "no such thing" setting. What I have done is
3846 to check that the total number of inodes is greater than zero before doing
3847 the test of how many are available.
3849 58. When a "warn" ACL statement has a log_message modifier, the message is
3850 remembered, and not repeated. This is to avoid a lot of repetition when a
3851 message has many recipients that cause the same warning to be written.
3852 Howewer, Exim was preserving the list of already written lines for an
3853 entire SMTP session, which doesn't seem right. The memory is now reset if a
3854 new message is started.
3856 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3857 debugging output unless log_rewrite was also set.
3859 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3860 of (int)(handle) when we know that handle contains (void *)(-1).
3862 61. The Exim daemon panic-logs an error return when it closes the incoming
3863 connection. However "connection reset by peer" seems to be common, and
3864 isn't really an error worthy of noting specially, so that particular error
3867 62. When Exim is trying to find all the local interfaces, it used to panic and
3868 die if the ioctl to get the interface flags failed. However, it seems that
3869 on at least one OS (Solaris 9) it is possible to have an interface that is
3870 included in the list of interfaces, but for which you get a failure error
3871 for this call. This happens when the interface is not "plumbed" into a
3872 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3873 failure of the "get flags" call assumes that the interface is down.
3875 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3876 makes life easier for people who are doing arithmetic on fields extracted
3877 from dates, where you often get leading zeros that should not be
3878 interpreted as octal.
3880 64. Added qualify_domain to the redirect router, to override the global
3883 65. If a pathologically long header line contained very many addresses (the
3884 report of this problem mentioned 10 000) and each of them was rewritten,
3885 Exim could use up a very large amount of memory. (It kept on making new
3886 copies of the header line as it rewrote, and never released the old ones.)
3887 At the expense of a bit more processing, the header rewriting function has
3888 been changed so that it no longer eats memory in this way.
3890 66. The generation of the Received: header has been moved from the time that a
3891 message starts to be received, to the time that it finishes. The timestamp
3892 in the Received: header should now be very close to that of the <= log
3893 line. There are two side-effects of this change:
3895 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3896 logged header lines no longer include the local Received: line, because
3897 it has not yet been created. The same applies to a copy of the message
3898 that is returned to a non-SMTP sender when a message is rejected.
3900 (b) When a filter file is tested using -bf, no additional Received: header
3901 is added to the test message. After some thought, I decided that this
3904 This change does not affect the value of $received_for. It is still set
3905 after address rewriting, but before local_scan() is called.
3907 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3909 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3910 gave an unhelpful panic error message, and a defer error. I have managed to
3911 change this behaviour so that it now rejects any supplied certificate,
3912 which seems right, as the list of acceptable certificates is empty.
3914 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3915 gave an unhelpful defer error. I have not managed to make this reject any
3916 supplied certificates, but the error message it gives is "no certificate
3917 supplied", which is not helpful.
3919 70. exigrep's output now also includes lines that are not associated with any
3920 message, but which match the given pattern. Implemented by a patch from
3921 Martin Sluka, which also tidied up the Perl a bit.
3923 71. Recipient callout verification, like sender verification, was using <> in
3924 the MAIL FROM command. This isn't really the right thing, since the actual
3925 sender may affect whether the remote host accepts the recipient or not. I
3926 have changed it to use the actual sender in the callout; this means that
3927 the cache record is now keyed on a recipient/sender pair, not just the
3928 recipient address. There doesn't seem to be a real danger of callout loops,
3929 since a callout by the remote host to check the sender would use <>.
3930 [SEE ABOVE: changed after hitting problems.]
3932 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3933 temporary errors. However, in the case of such a code being given after
3934 the end of a data transmission (i.e. after ".") Exim was failing to write
3935 a retry record for the message. (Yes, there was some broken host that was
3936 actually sending 8xx at this point.)
3938 73. An unknown lookup type in a host list could cause Exim to panic-die when
3939 the list was checked. (An example that provoked this was putting <; in the
3940 middle of a list instead of at the start.) If this happened during a DATA
3941 ACL check, a -D file could be left lying around. This kind of configuration
3942 error no longer causes Exim to die; instead it causes a defer errror. The
3943 incident is still logged to the main and panic logs.
3945 74. Buglet left over from Exim 3 conversion. The message "too many messages
3946 in one connection" was written to the rejectlog but not the mainlog, except
3947 when address rewriting (yes!) was being logged.
3949 75. Added write_rejectlog option.
3951 76. When a system filter was run not as root (that is, when system_filter_user
3952 was set), the values of the $n variables were not being returned to the
3953 main process; thus, they were not subsequently available in the $sn
3956 77. Added +return_path_on_delivery log selector.
3958 78. A connection timeout was being treated differently from recipients deferred
3959 when testing hosts_max_try with a message that was older than the host's
3960 retry timeout. (The host should not be counted, thus allowing all hosts to
3961 be tried at least once before bouncing.) This may have been the cause of an
3962 occasionally reported bug whereby a message would remain on the queue
3963 longer than the retry timeout, but would be bounced if a delivery was
3964 forced. I say "may" because I never totally pinned down the problem;
3965 setting up timeout/retry tests is difficult. See also the next item.
3967 79. The ultimate address timeout was not being applied to errors that involved
3968 a combination of host plus message (for example, a timeout on a MAIL
3969 command). When an address resolved to a number of possible hosts, and they
3970 were not all tried for each delivery (e.g. because of hosts_max_try), a
3971 message could remain on the queue longer than the retry timeout.
3973 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3976 81. Fixed an obscure SMTP outgoing bug which required at least the following
3977 conditions: (a) there was another message waiting for the same server;
3978 (b) the server returned 5xx to all RCPT commands in the first message so
3979 that the message was not completed; (c) the server dropped the connection
3980 or gave a negative response to the RSET that Exim sends to abort the
3981 transaction. The observed case was a dropped connection after DATA that had
3982 been sent in pipelining mode. That is, the server had advertised PIPELINING
3983 but was not implementing it correctly. The effect of the bug was incorrect
3984 behaviour, such as trying another host, and this could lead to a crash.
3990 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3991 and daemon.c were passed as pointers to ints; they should have been
3992 pointers to socklen_t variables (which are typically unsigned ints).
3994 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3997 3. Fixed a really odd bug that affected only the testing scheme; patching a
3998 certain fixed string in the binary changed the value of another string that
3999 happened to be identical to the end of the original first string.
4001 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4002 name", it returns that address as the IP address. On some operating
4003 systems (e.g. Solaris), it also passes back the IP address string as the
4004 "host name". However, on others (e.g. Linux), it passes back an empty
4005 string. Exim wasn't checking for this, and was changing the host name to an
4006 empty string, assuming it had been canonicized.
4008 5. Although rare, it is permitted to have more than one PTR record for a given
4009 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4010 all the names associated with an address, because they do in Solaris.
4011 However, it seems that they do not in Linux for data that comes from the
4012 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4013 I found this out when I moved to a new Linux workstation and tried to run
4014 the Exim test suite.
4016 To get round this problem I have changed the code so that it now does its
4017 own call to the DNS to look up PTR records when searching for a host name.
4018 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4019 addresses that are only in /etc/hosts are still found.
4021 This behaviour is, however, controlled by an option called host_lookup_
4022 order, which defaults to "bydns:byaddr". If people want to use the other
4023 order, or indeed, just use one or the other means of lookup, they can
4024 specify it in this variable.
4026 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4027 some operating systems, this comes back from gethostbyaddr() as an empty
4028 string, and this is what Exim used to test for. However, it seems that in
4029 other systems, "." is yielded. Exim now tests for this case too.
4031 7. The values of check_spool_space and check_log_space are now held internally
4032 as a number of kilobytes instead of an absolute number of bytes. If a
4033 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4034 kilobyte. This means that much larger values can be stored.
4036 8. Exim monitor: an attempt to get the action menu when not actually pointing
4037 at a message produces an empty menu entitled "No message selected". This
4038 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4039 no entries in it ("Shell widget menu has zero width and/or height"). So I
4040 have added a single, blank menu entry in this case.
4042 9. Added ${quote_local_part.
4044 10. MIME decoding is now applied to the contents of Subject: header lines when
4047 11. Now that a reference to $sender_host_address automatically causes a reverse
4048 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4049 host lookup before query-style lookups in lists that might use this
4050 variable. This has therefore been abolished, and the "net-" prefix is no
4051 longer necessary for query-style lookups.
4053 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4054 have been a typo for LFLAGS, so it has been changed.
4056 13. The install script calls Exim with "-C /dev/null" in order to find the
4057 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4058 to be output. Howeve, since Exim outputs its version number before the
4059 error, it didn't break the script. It just looked ugly. I fixed this by
4060 always allowing "-C /dev/null" if the caller is root.
4062 14. Ignore overlarge ACL variable number when reading spool file - insurance
4063 against a later release with more variables having written the file.
4065 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4066 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4067 this, as well as the form without the "IPv6" on the front.
4069 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4070 OS/Makefile-Darwin file.
4072 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4073 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4075 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4076 variable. However, some OS use ints for inodes. I've added cast to long int
4077 to get rid of the compiler warning.
4079 19. I had forgotten to lock out "/../" in configuration file names when
4080 ALT_CONFIG_PREFIX was set.
4082 20. Routers used for verification do not need to specify transports. However,
4083 if such a router generated a host list, and callout was configured, Exim
4084 crashed, because it could not find a port number from the (non-existent)
4085 transport. It now assumes port 25 in this circumstance.
4087 21. Added the -t option to exigrep.
4089 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4090 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4091 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4092 them is set, LOOKUP_LSEARCH is forced.
4094 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4096 24. Added sender and host information to the "rejected by local_scan()" log
4097 line; previously there was no indication of these.
4099 25. Added .include_if_exists.
4101 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4102 Linux. It turns out that not all file systems support this. Apparently some
4103 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4104 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4105 not supported on the file descriptor, is now ignored when Exim is trying to
4106 sync a directory. This applies only to Linux.
4108 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4110 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4111 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4114 29. Added support for maildirsize files from supplied patch (modified a bit).
4116 30. The use of :fail: followed by an empty string could lead Exim to respond to
4117 sender verification failures with (e.g.):
4119 550 Verification failed for <xxx>
4120 550 Sender verify failed
4122 where the first response line was missing the '-' that indicates it is not
4123 the final line of the response.
4125 31. The loop for finding the name of the user that called Exim had a hardwired
4126 limit of 10; it now uses the value of finduser_retries, which is used for
4127 all other user lookups.
4129 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4132 33. Exim was neglecting to zero errno before one call of strtol() when
4133 expanding a string and expecting an integer value. On some systems this
4134 resulted in spurious "integer overflow" errors. Also, it was casting the
4135 result into an int without checking.
4137 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4138 did not work. The code looks as if it has *never* worked, though it appears
4139 to have been documented since at least releast 1.62. I have made it work.
4141 35. The "timeout_DNS" error in retry rules, also documented since at least
4142 1.62, also never worked. As it isn't clear exactly what this means, and
4143 clearly it isn't a major issue, I have abolished the feature by treating it
4144 as "timeout", and writing a warning to the main and panic logs.
4146 36. The display of retry rules for -brt wasn't always showing the error code
4149 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4150 timeout_connect_A, timeout_connect_MX.
4152 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4153 to the empty sender.
4155 39. The daemon was not analysing the content of -oX till after it had closed
4156 stderr and disconnected from the controlling terminal. This meant that any
4157 syntax errors were only noted on the panic log, and the return code from
4158 the command was 0. By re-arranging the code a little, I've made the
4159 decoding happen first, so such errors now appear on stderr, and the return
4160 code is 1. However, the actual setting up of the sockets still happens in
4161 the disconnected process, so errors there are still only recorded on the
4164 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4165 connections (as happens on some IP stacks) was logged at start up time as
4166 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4167 it from "IPv6 and IPv4", which means that two separate sockets are being
4170 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4171 says whether AF_INET or AF_INET6 was passed as an argument.
4173 42. Exiwhat output was messed up when time zones were included in log
4176 43. Exiwhat now gives more information about the daemon's listening ports,
4177 and whether -tls-on-connect was used.
4179 44. The "port" option of the smtp transport is now expanded.
4181 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4182 silently ignored. Now an error message is written to the main and panic
4185 46. There's a new ACL modifier called "logwrite" which writes to a log file
4186 as soon as it is encountered.
4188 47. Added $local_user_uid and $local_user_gid at routing time.
4190 48. Exim crashed when trying to verify a sender address that was being
4193 49. Exim was recognizing only a space character after ".include". It now also
4194 recognizes a tab character.
4196 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4197 extracting the relevant information from the specification. The man page no
4198 longer contains scrambled data for the -d option, and I've added a section
4199 at the front about calling Exim under different names.
4201 51. Added "extra_headers" argument to the "mail" command in filter files.
4203 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4206 53. Installed eximstats 1.29.
4208 54. Added transport_filter_timeout as a generic transport option.
4210 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4211 Cc: header lines. This was required by RFC 822, but it not required by RFC
4214 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4215 incoming messages that did not have them. Now it does so only if the
4216 message originates locally, that is, if there is no associated remote host
4217 address. When Resent- header lines are present, this applies to the Resent-
4218 lines rather than the non-Resent- lines.
4220 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4221 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4223 58. Messages for configuration errors now include the name of the main
4224 configuration file - useful now that there may be more than one file in a
4225 list (.included file names were always shown).
4227 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4228 for those rare installations that do not start the daemon as root or run it
4229 setuid root. I've cut out the call to initgroups() if the daemon is not
4232 60. The Exim user and group can now be bound into the binary as text strings
4233 that are looked up at the start of Exim's processing.
4235 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4237 62. Added $mailstore_basename variable.
4239 63. Installed patch to sieve.c from Michael Haardt.
4241 64. When Exim failed to open the panic log after failing to open the main log,
4242 the original message it was trying to log was written to stderr and debug
4243 output, but if they were not available (the usual case in production), it
4244 was lost. Now it is written to syslog before the two lines that record the
4245 failures to open the logs.
4247 65. Users' Exim filters run in subprocesses under the user's uid. It is
4248 possible for a "deliver" command or an alias in a "personal" command to
4249 provoke an address rewrite. If logging of address rewriting is configured,
4250 this fails because the process is not running as root or exim. There may be
4251 a better way of dealing with this, but for the moment (because 4.30 needs
4252 to be released), I have disabled address rewrite logging when running a
4253 filter in a non-root, non-exim process.
4259 1. The buildconfig auxiliary program wasn't quoting the value set for
4260 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4261 not defined. This bug was masked in 4.22 by the effect that was fixed in
4264 2. Some messages that were rejected after a message id was allocated were
4265 shown as "incomplete" by exigrep. It no longer does this for messages that
4266 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4268 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4269 have allow_domain_literals set, the ID did not get logged in the <= line.
4270 Domain literals are now always recognized in Message-ID: header lines.
4272 4. The first argument for a ${extract expansion item is the key name or field
4273 number. Leading and trailing spaces in this item were not being ignored,
4274 causing some misleading effects.
4276 5. When deliver_drop_privilege was set, single queue runner processes started
4277 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4278 same command in the process it spins off) were not dropping privilege.
4280 6. When the daemon running as "exim" started a queue runner, it always
4281 re-executed Exim in the spun-off process. This is a waste of effort when
4282 deliver_drop_privilege is set. The new process now just calls the
4283 queue-runner function directly.
4289 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4292 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4293 ignored. Though the use of -oP was forcing the writing of a pid file, it
4294 was always written to the default place.
4296 3. If the message "no IP address found for host xxxx" is generated during
4297 incoming verification, it is now followed by identification of the incoming
4298 connection (so you can more easily find what provoked it).
4300 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4302 5. Added some features to "harden" Exim a bit more against certain attacks:
4304 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4305 be put in Local/Makefile. This is like the never_users runtime option,
4306 but it cannot be overridden. The default setting is "root".
4308 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4309 prefix string with which any file named in a -C command line option
4312 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4313 is retained for -C and -D only if the caller of Exim is root. Without
4314 it, the exim user may also use -C and -D and retain privilege.
4316 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4317 command line option is disabled.
4319 6. Macro names set by the -D option must start with an upper case letter, just
4320 like macro names defined in the configuration file.
4322 7. Added "dereference=" facility to LDAP.
4324 8. Two instances of the typo "uknown" in the source files are fixed.
4326 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4327 the Configure-Makefile script screwed up while processing it.
4329 10. Incorporated PCRE 4.4.
4331 11. The SMTP synchronization check was not operating right at the start of an
4332 SMTP session. For example, it could not catch a HELO sent before the client
4333 waited for the greeting. There is now a check for outstanding input at the
4334 point when the greeting is written. Because of the duplex, asynchronous
4335 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4336 way, but not yet received, when the check is performed.
4338 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4339 on TCP/IP sockets, because this apparently causes some broken clients to
4342 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4343 unchanged) from the Cygwin maintainer.
4345 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4346 was supported instead of testing for mbx. Effectively a typo.
4348 15. The spa authenticator server code was not checking that the input it
4349 received was valid base64.
4351 16. The debug output line for the "set" modifier in ACLs was not showing the
4352 name of the variable that was being set.
4354 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4356 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4357 lookup on its own. Something else was needed to trigger the lookup. For
4358 example, a match in host_lookup or the need for a host name in a host list.
4359 Now, if $sender_host_name is referenced and the host name has not yet been
4360 looked up, a lookup is performed. If the lookup fails, the variable remains
4361 empty, and $host_lookup_failed is set to "1".
4363 19. Added "eqi" as a case-independent comparison operator.
4365 20. The saslauthd authentication condition could segfault if neither service
4366 nor realm was specified.
4368 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4369 error message that was logged was misleading, and incoming SMTP
4370 connections were dropped. The message is now more accurate, and temporary
4371 errors are given to SMTP connections.
4373 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4374 (except RSET) with a 421 error, until QUIT is received. However, it was
4375 failing to send a response to QUIT.
4377 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4378 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4379 ACL is now run after the helo_try_verify_hosts code.
4381 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4382 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4383 case-independent, so other case variants are also recognized). Apparently
4384 some systems use these upper case variants.
4386 25. If more than two messages were waiting for the same host, and a transport
4387 filter was specified for the transport, Exim sent two messages over the
4388 same TCP/IP connection, and then failed with "socket operation on non-
4389 socket" when it tried to send the third.
4391 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4393 27. The extern definition of crypt16() in expand.c was not being excluded when
4394 the OS had its own crypt16() function.
4396 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4397 as a preferred synonym for return_size_limit, both as an option and as an
4400 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4402 30. Changed the default configuration ACL to relax the local part checking rule
4403 for addresses that are not in any local domains. For these addresses,
4404 slashes and pipe symbols are allowed within local parts, but the sequence
4405 /../ is explicitly forbidden.
4407 31. SPA server authentication was not clearing the challenge buffer before
4410 32. log_message in a "warn" ACL statement was writing to the reject log as
4411 well as to the main log, which contradicts the documentation and doesn't
4412 seem right (because no rejection is happening). So I have stopped it.
4414 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4415 However, I am unable to do any testing of this.
4417 34. Fixed an infelicity in the appendfile transport. When checking directories
4418 for a mailbox, to see if any needed to be created, it was accidentally
4419 using path names with one or more superfluous leading slashes; tracing
4420 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4422 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4423 contents are added to the log line that is written for every discarded
4424 recipient. (Previously a log_message setting was ignored.)
4426 36. The ${quote: operator now quotes the string if it is empty.
4428 37. The install script runs exim in order to find its version number. If for
4429 some reason other than non-existence or emptiness, which it checks, it
4430 could not run './exim', it was installing it with an empty version number,
4431 i.e. as "exim-". This error state is now caught, and the installation is
4434 38. An argument was missing from the function that creates an error message
4435 when Exim fails to connect to the socket for saslauthd authentication.
4436 This could cause Exim to crash, or give a corrupted message.
4438 39. Added isip, isip4, and isip6 to ${if conditions.
4440 40. The ACL variables $acl_xx are now saved with the message, and can be
4441 accessed later in routers, transports, and filters.
4443 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4444 strings in the file are not string-expanded.
4446 42. If a MAIL command specified a SIZE value that was too large to fit into an
4447 int variable, the check against message_size_limit failed. Such values are
4448 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4449 day this will have to be increased, but I don't think I want to be around
4450 when emails are that large.
4457 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4458 iconv() is not standard in FreeBSD.
4460 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4461 IPv6 enabled. The observed symptom was a segmentation fault on return from
4462 the function os_common_find_running_interfaces() in src/os.c.
4464 3. In the check_special_case() function in daemon.c I had used "errno" as an
4465 argument name, which causes warnings on some systems. This was basically a
4466 typo, since it was named "eno" in the comments!
4468 4. The code that waits for the clock to tick (at a resolution of some fraction
4469 of a second) so as to ensure message-id uniqueness was always waiting for
4470 at least one whole tick, when it could have waited for less. [This is
4471 almost certainly not relevant at current processor speeds, where it is
4472 unlikely to ever wait at all. But we try to future-proof.]
4474 5. The function that sleeps for a time interval that includes fractions of a
4475 second contained a race. It did not block SIGALRM between setting the
4476 timer, and suspending (a couple of lines later). If the interval was short
4477 and the sigsuspend() was delayed until after it had expired, the suspension
4478 never ended. On busy systems this could lead to processes getting stuck for
4481 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4482 process, before it forks any delivery processes. The open lookup caching
4483 mechanism meant that the open file or database connection was passed into
4484 the delivery process. The problem was that delivery processes always tidy
4485 up cached lookup data. This could cause a problem for the next delivery
4486 process started by the queue runner, because the external queue runner
4487 process does not know about the closure. So the next delivery process
4488 still has data in the lookup cache. In the case of a file lookup, there was
4489 no problem because closing a file descriptor in a subprocess doesn't affect
4490 the parent. However, if the lookup was caching a connection to a database,
4491 the connection was closed, and the second delivery process was likely to
4492 see errors such as "PGSQL: query failed: server closed the connection
4493 unexpectedly". The problem has been fixed by closing all cached lookups
4494 in a queue runner before running a delivery process.
4496 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4497 seem to have the "const" qualifier which it has on other OS. I've
4500 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4501 *of the same type* (client or server) with the same public name that an
4502 error should be diagnosed.
4504 9. When Exim looked up a host name for an IP address, but failed to find the
4505 original IP address when looking up the host name (a safety check), it
4506 output the message "<ip address> does not match any IP for NULL", which was
4507 confusing, to say the least. The bug was that the host name should have
4508 appeared instead of "NULL".
4510 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4511 user that is built into the binary, and the -C or -D options is used, root
4512 privilege is dropped before the configuration file is read. In addition,
4513 logging is switched to stderr instead of the normal log files. If the
4514 configuration then re-defines the Exim user, the unprivileged environment
4515 is probably not what is expected, so Exim logs a panic warning message (but
4518 However, if deliver_drop_privilege is set, the unprivileged state may well
4519 be exactly what is intended, so the warning has been cut out in that case,
4520 and Exim is allowed to try to write to its normal log files.
4526 1. smtp_return_error_details was not giving details for temporary sender
4527 or receiver verification errors.
4529 2. Diagnose a configuration error if two authenticators have the same public
4532 3. Exim used not to create the message log file for a message until the first
4533 delivery attempt. This could be confusing when incoming messages were held
4534 for policy or load reasons. The message log file is now created at the time
4535 the message is received, and an initial "Received" line is written to it.
4537 4. The automatically generated man page for command line options had a minor
4538 bug that caused no ill effects; however, a more serious problem was that
4539 the procedure for building the man page automatically didn't always
4540 operate. Consequently, release 4.20 contains an out-of-date version. This
4541 shouldn't happen again.
4543 5. When building Exim with embedded Perl support, the script that builds the
4544 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4545 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4547 6. The freeze_tell option was not being used for messages that were frozen on
4548 arrival, either by an ACL or by local_scan().
4550 7. Added the smtp_incomplete_transaction log selector.
4552 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4553 was accepting AUTH without a new EHLO.
4555 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4556 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4559 10. Logging of TCP/IP connections (when configured) now happens in the main
4560 daemon process instead of the child process, so that the TCP/IP connection
4561 count is more accurate (but it can never be perfect).
4563 11. The use of "drop" in a nested ACL was not being handled correctly in the
4564 outer ACL. Now, if condition failure induced by the nested "drop" causes
4565 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4566 or "require"), the connection is dropped.
4568 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4569 that yield "discard" can now be used with an "accept" or a "discard" verb,
4570 but an error is generated for any others (because I can't see a useful way
4571 to define what should happen).
4573 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4574 are now processed in the same way as lines in the Exim configuration file.
4575 In particular, continuation lines are supported.
4577 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4579 15. Added -ti meaning -t -i.
4581 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4582 domains, and warn by logging if others are found.
4584 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4585 returned from GIFCONF when Exim is trying to find the list of interfaces on
4586 a host. The code in os.c has been modified to copy each ifreq to an aligned
4587 structure in all cases.
4589 Also, in some cases, the returned ifreq's were being copied to a 'struct
4590 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4591 means the last couple of bytes of an IPv6 address could be chopped if the
4592 ifreq contained only a normal sockaddr (14 bytes storage).
4594 18. Named domain lists were not supported in the hosts_treat_as_local option.
4595 An entry such as +xxxx was not recognized, and was treated as a literal
4598 19. Ensure that header lines added by a DATA ACL are included in the reject log
4599 if the ACL subsequently rejects the message.
4601 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4602 MD5 (which is deprecated).
4604 21. When testing a filter file using -bf, Exim was writing a message when it
4605 took the sender from a "From " line in the message, but it was not doing so
4606 when it took $return_path from a Return-Path: header line. It now does.
4608 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4609 with a valid header line field name (a series of printing characters
4610 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4612 23. Changed "disc" in the source to "disk" to conform to the documentation and
4613 the book and for uniformity.
4615 24. Ignore Sendmail's -Ooption=value command line item.
4617 25. When execve() failed while trying to run a command in a pipe transport,
4618 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4619 could be confused with a return value of 69 from the command itself. This
4620 has been changed to 127, the value the shell returns if it is asked to run
4621 a non-existent command. The wording for the related log line suggests a
4622 non-existent command as the problem.
4624 26. If received_header_text expands to an empty string, do not add a Received:
4625 header line to the message. (Well, it adds a token one on the spool, but
4626 marks it "old" so that it doesn't get used or transmitted.)
4628 27. Installed eximstats 1.28 (addition of -nt option).
4630 28. There was no check for failure on the call to getsockname() in the daemon
4631 code. This can fail if there is a shortage of resources on the system, with
4632 ENOMEM, for example. A temporary error is now given on failure.
4634 29. Contrary to the C standard, it seems that in some environments, the
4635 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4636 program. Exim now does this explicitly; it affects the formatting of
4637 timestamps using strftime().
4639 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4640 complaints. I've now initialized all the variables, to avoid this.
4642 32. Header lines added by a system filter were not being "seen" during
4643 transport-time rewrites.
4645 33. The info_callback() function passed to OpenSSL is set up with type void
4646 (*)(SSL *, int, int), as described somewhere. However, when calling the
4647 function (actually a macro) that sets it up, the type void(*)() is
4648 expected. I've put in a cast to prevent warnings from picky compilers.
4650 34. If a DNS black list lookup found a CNAME record, but there were no A
4651 records associated with the domain it pointed at, Exim crashed.
4653 35. If a DNS black list lookup returned more than one A record, Exim ignored
4654 all but the first. It now scans all returned addresses if a particular IP
4655 value is being sought. In this situation, the contents of the
4656 $dnslist_value variable are a list of all the addresses, separated by a
4659 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4660 to accept a host name and all its aliases if the forward lookup for any of
4661 them yielded the IP address of the incoming connection. Now it accepts only
4662 those names whose forward lookup yields the correct IP address. Any other
4663 names are discarded. This closes a loophole whereby a rogue DNS
4664 administrator could create reverse DNS records to break through a
4665 wildcarded host restriction in an ACL.
4667 37. If a user filter or a system filter that ran in a subprocess used any of
4668 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4669 the wrong values were passed to the pipe command ($thisaddress had the
4670 value of $0, $0 had the value of $1, etc). This bug was introduced by
4671 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4673 38. Improved the line breaking for long SMTP error messages from ACLs.
4674 Previously, if there was no break point between 40 and 75 characters, Exim
4675 left the rest of the message alone. Two changes have been made: (a) I've
4676 reduced the minimum length to 35 characters; (b) if it can't find a break
4677 point between 35 and 75 characters, it looks ahead and uses the first one
4678 that it finds. This may give the occasional overlong line, but at least the
4679 remaining text gets split now.
4681 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4682 file descriptors might be low, and that setting 1000 would always raise it.
4683 It turns out that in some environments, the limit is already over 1000 and
4684 that lowering it causes trouble. So now Exim takes care not to decrease it.
4686 40. When delivering a message, the value of $return_path is set to $sender_
4687 address at the start of routing (routers may change the value). By an
4688 oversight, this default was not being set up when an address was tested by
4689 -bt or -bv, which affected the outcome if any router or filter referred to
4692 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4693 log only when "message" or "log_message" are set. However, if one of the
4694 conditions was an address verification, or a call to a nested ACL, the
4695 messages generated by the underlying test were being passed through. This
4696 no longer happens. The underlying message is available in $acl_verify_
4697 message for both "message" and "log_message" expansions, so it can be
4698 passed through if needed.
4700 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4701 new expansion $bh_ to give the encoded byte string without charset
4702 translation. Translation happens only if iconv() is available; HAVE_ICONV
4703 indicates this at build time. HEADERS_CHARSET gives the charset to
4704 translate to; headers_charset can change it in the configuration, and
4705 "headers charset" can change it in an individual filter file.
4707 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4708 that creates RFC 2047 encoded "words" labels them as that charset instead
4709 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4710 expansion operator; (ii) when Exim creates a From: line for a local
4711 message; (iii) when a header line is rewritten to include a "phrase" part.
4713 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4714 buggy, causing it to skip the first lines of messages whose message ID
4715 ended in 'D'. This would not have bitten before Exim release 4.14, because
4716 message IDs were unlikely to end in 'D' before then. The effect was to have
4717 incorrect size information for certain domains.
4719 45. #include "config.h" was missing at the start of the crypt16.c module. This
4720 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4723 46. If there was a timeout during a "random" callout check, Exim treated it as
4724 a failure of the random address, and carried on sending RSET and the real
4725 address. If the delay was just some slowness somewhere, the response to the
4726 original RCPT would be taken as a response to RSET and so on, causing
4727 mayhem of various kinds.
4729 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4730 when I implemented it. It didn't allow for the fact that some option values
4731 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4732 the right test for positive values.
4734 48. Domain names in DNS records are case-independent. Exim always looks them up
4735 in lower case. Some resolvers return domain names in exactly the case they
4736 appear in the zone file, that is, they may contain uppercase letters. Not
4737 all resolvers do this - some return always lower case. Exim was treating a
4738 change of case by a resolver as a change of domain, similar to a widening
4739 of a domain abbreviation. This triggered its re-routing code and so it was
4740 trying to route what was effectively the same domain again. This normally
4741 caused routing to fail (because the router wouldn't handle the domain
4742 twice). Now Exim checks for this case specially, and just changes the
4743 casing of the domain that it ultimately uses when it transmits the message
4746 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4749 50. If a filter generated a file delivery with a non-absolute name (possible if
4750 no home directory exists for the router), the forbid_file option was not
4753 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4754 the existing equality matching.
4756 52. Exim was using ints instead of ino_t variables in some places where it was
4757 dealing with inode numbers.
4759 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4760 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4761 called TMPDIR, and if it finds it is different, it changes its value.
4763 54. The smtp_printf() function is now made available to local_scan() so
4764 additional output lines can be written before returning. There is also an
4765 smtp_fflush() function to enable the detection of a dropped connection.
4766 The variables smtp_input and smtp_batched_input are exported to
4769 55. Changed the default runtime configuration: the message "Unknown user"
4770 has been removed from the ACL, and instead placed on the localuser router,
4771 using the cannot_route_message feature. This means that any verification
4772 failures that generate their own messages won't get overridden. Similarly,
4773 the "Unrouteable address" message that was in the ACL for unverifiable
4774 relay addresses has also been removed.
4776 56. Added hosts_avoid_esmtp to the smtp transport.
4778 57. The exicyclog script was not checking for the esoteric option
4779 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4780 will work only if exicyclog is run under the appropriate euid.
4782 58. Following a discussion on the list, the rules by which Exim recognises line
4783 endings on incoming messages have been changed. The -dropcr and drop_cr
4784 options are now no-ops, retained only for backwards compatibility. The
4785 following line terminators are recognized: LF CRLF CR. However, special
4786 processing applies to CR:
4788 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4789 nor a local message in the state where . is a terminator.
4791 (ii) If a bare CR is encountered in a header line, an extra space is added
4792 after the line terminator so as not to end the header. The reasoning
4793 behind this is that bare CRs in header lines are most likely either
4794 to be mistakes, or people trying to play silly games.
4796 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4797 was being incorrectly given as 18 bytes larger than it should have been.
4798 This is a VOB (very old bug).
4800 60. This may never have affected anything current, but just in case it has:
4801 When the local host is found other than at the start of a list of hosts,
4802 the local host, those with the same MX, and any that follow, are discarded.
4803 When the list in question was part of a longer list of hosts, the following
4804 hosts (not currently being processed) were also being discarded. This no
4805 longer happens. I'm not sure if this situation could ever has previously
4808 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4811 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4812 header. This is something that is recommended in a new Internet Draft, and
4813 is something that is documented as being done by Sendmail. There are two
4814 possible values. For messages generated by the autoreply transport, Exim
4817 Auto-Submitted: auto-replied
4819 whereas for all other generated messages (e.g. bounces) it adds
4821 Auto-Submitted: auto-generated
4823 63. The "personal" condition in filters now includes a test for the
4824 Auto-Submitted: header. If it contains the string "auto-" the message it
4825 not considered personal.
4827 64. Added rcpt_include_affixes as a generic transport option.
4829 65. Added queue_only_override (default true).
4831 66. Added the syslog_duplication option.
4833 67. If what should have been the first header line of a message consisted of
4834 a space followed by a colon, Exim was mis-interpreting it as a header line.
4835 It isn't of course - it is syntactically invalid and should therefore be
4836 treated as the start of the message body. The misbehaviour could have
4837 caused a number of strange effects, including loss of data in subsequent
4838 header lines, and spool format errors.
4840 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4841 client host had authenticated. This control can now be exercised by an ACL
4842 for more flexibility.
4844 69. By default, callouts do not happen when testing with -bh. There is now a
4845 variant, -bhc, which does actually run the callout code, including
4846 consulting and updating the callout cache.
4848 70. Added support for saslauthd authentication, courtesy of Alexander
4851 71. If statvfs() failed on the spool or log directories while checking their
4852 size for availability, Exim confusingly gave the error "space shortage".
4853 Furthermore, in debugging mode it crashed with a floating point exception.
4854 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4855 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4856 serious problem, Exim now writes to the main and panic logs when this
4857 happens, with details of the failure. It then refuses to accept the
4858 incoming message, giving the message "spool directory problem" or "log
4859 directory problem" with a 421 code for SMTP messages.
4861 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4862 0, 1, and 2 exist, because some OS complain for execs without them (see
4863 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4864 descriptors. However, the code omitted to check that the open succeeded,
4865 causing mysterious errors if for some reason the permissions on /dev/null
4866 got screwed. Now Exim writes a message to the main and panic logs, and
4867 bombs out if it can't open /dev/null.
4869 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4870 interact so that it is all more flexible. It is supposed to remain
4871 backwards compatible. Also added extra_local_interfaces.
4873 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4874 to bomb out with an assertion failure - to the client this appears as a
4875 connection drop. This problem occurs in the part of the code that was taken
4876 from the Samba project. Fortunately, the assertion is in a very simple
4877 function, so I have fixed this by reproducing the function inline in the
4878 one place where it is called, and arranging for authentication to fail
4879 instead of killing the process with assert().
4881 75. The SPA client code was not working when the server requested OEM rather
4882 than Unicode encoding.
4884 76. Added code to make require_files with a specific uid setting more usable in
4885 the case where statting the file as root fails - usually a non-root-mounted
4886 NFS file system. When this happens and the failure is EACCES, Exim now
4887 forks a subprocess and does the per-uid checking as the relevant uid.
4889 77. Added process_log_path.
4891 78. If log_file_path was not explicitly set, a setting of check_log_space or
4892 check_log_inodes was ignored.
4894 79. If a space check for the spool or log partitions fails, the incident is now
4895 logged. Of course, in the latter case the data may get lost...
4897 80. Added the %p formatting code to string_format() so that it can be used to
4898 print addresses in debug_print(). Adjusted all the address printing in the
4899 debugging in store.c to use %p rather than %d.
4901 81. There was a concern that a line of code in smtp_in.c could overflow a
4902 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4903 initially expressed, the concern was not well-founded, because trailing
4904 spaces are removed early. However, if the trailing spaces were followed by
4905 a NULL, they did not get removed, so the overflow was possible. Two fixes
4908 (a) I re-wrote the offending code in a cleaner fashion.
4909 (b) If an incoming SMTP command contains a NULL character, it is rejected
4912 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4913 runs initgroups(), so that if the Exim user is in any additional groups,
4914 they will be used during message reception.
4920 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob