1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.569 2009/10/14 14:48:41 nm4 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: Improvement to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
57 TF/05 Leading white space used to be stripped from $spam_report which
58 wrecked the formatting. Now it is preserved.
60 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
61 that they are available at delivery time.
63 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
65 TF/08 TLS error reporting now respects the incoming_interface and
66 incoming_port log selectors.
68 TF/09 Produce a more useful error message if an SMTP transport's hosts
69 setting expands to an empty string.
71 NM/06 Bugzilla 744: EXPN did not work under TLS.
72 Patch provided by Phil Pennock
74 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
75 Patch provided by Richard Godbee
77 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
78 acl_smtp_notquit, added index entry.
80 NM/09 Bugzilla 787: Potential buffer overflow in string_format
81 Patch provided by Eugene Bujak
83 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
84 Patch provided by Maxim Dounin
86 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
87 Patch provided by Phil Pennock
89 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
91 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
92 Patch provided by Brad "anomie" Jorsch
94 NM/14 Bugzilla 614: Added timeout setting to transport filter.
95 Patch provided by Dean Brooks
97 TK/05 Add native DKIM support (does not depend on external libraries).
99 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
100 Patch provided by Graeme Fowler
102 NM/16 Bugzilla 851: Documentation example syntax fix.
104 NM/17 Changed NOTICE file to remove references to embedded PCRE.
106 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
108 NM/19 Bugzilla 745: TLS version reporting
109 Patch provided by Phil Pennock
111 NM/20 Bugzilla 167: bool: condition support
112 Patch provided by Phil Pennock
118 TK/01 Add preliminary DKIM support. Currently requires a forked version of
119 ALT-N's libdkim that I have put here:
120 http://duncanthrax.net/exim-experimental/
122 Note to Michael Haardt: I had to rename some vars in sieve.c. They
123 were called 'true' and it seems that C99 defines that as a reserved
124 keyword to be used with 'bool' variable types. That means you could
125 not include C99-style headers which use bools without triggering
126 build errors in sieve.c.
128 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
129 as mailq or other aliases. Changed the --help handling significantly
130 to do whats expected. exim_usage() emits usage/help information.
132 SC/01 Added the -bylocaldomain option to eximstats.
134 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
136 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
138 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
144 PH/01 Another patch from the Sieve maintainer.
146 PH/02 When an IPv6 address is converted to a string for single-key lookup
147 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
148 dots are used instead of colons so that keys in lsearch files need not
149 contain colons. This was done some time before quoting was made available
150 in lsearch files. However, iplsearch files do require colons in IPv6 keys
151 (notated using the quote facility) so as to distinguish them from IPv4
152 keys. This meant that lookups for IP addresses in host lists did not work
153 for iplsearch lookups.
155 This has been fixed by arranging for IPv6 addresses to be expressed with
156 colons if the lookup type is iplsearch. This is not incompatible, because
157 previously such lookups could never work.
159 The situation is now rather anomolous, since one *can* have colons in
160 ordinary lsearch keys. However, making the change in all cases is
161 incompatible and would probably break a number of configurations.
163 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
166 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
167 conversion specification without a maximum field width, thereby enabling
168 a rogue spamd server to cause a buffer overflow. While nobody in their
169 right mind would setup Exim to query an untrusted spamd server, an
170 attacker that gains access to a server running spamd could potentially
171 exploit this vulnerability to run arbitrary code as the Exim user.
173 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
174 $primary_hostname instead of what libspf2 thinks the hosts name is.
176 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
177 a directory entry by the name of the lookup key. Previously, if a
178 symlink pointed to a non-existing file or a file in a directory that
179 Exim lacked permissions to read, a lookup for a key matching that
180 symlink would fail. Now it is enough that a matching directory entry
181 exists, symlink or not. (Bugzilla 503.)
183 PH/03 The body_linecount and body_zerocount variables are now exported in the
186 PH/04 Added the $dnslist_matched variable.
188 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
189 This means they are set thereafter only if the connection becomes
192 PH/06 Added the client_condition to authenticators so that some can be skipped
193 by clients under certain conditions.
195 PH/07 The error message for a badly-placed control=no_multiline_responses left
196 "_responses" off the end of the name.
198 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
200 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
201 (without spaces) instead of just copying the configuration text.
203 PH/10 Added the /noupdate option to the ratelimit ACL condition.
205 PH/11 Added $max_received_linelength.
207 PH/12 Added +ignore_defer and +include_defer to host lists.
209 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
210 way in which PCRE > 7.0 is built.
212 PH/14 Implemented queue_only_load_latch.
214 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
215 MAIL command. The effect was to mangle the value on 64-bit systems.
217 PH/16 Another patch from the Sieve maintainer.
219 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
221 PH/18 If a system quota error occurred while trying to create the file for
222 a maildir delivery, the message "Mailbox is full" was not appended to the
223 bounce if the delivery eventually timed out. Change 4.67/27 below applied
224 only to a quota excession during the actual writing of the file.
226 PH/19 It seems that peer DN values may contain newlines (and other non-printing
227 characters?) which causes problems in log lines. The DN values are now
228 passed through string_printing() before being added to log lines.
230 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
231 and InterBase are left for another time.)
233 PH/21 Added message_body_newlines option.
235 PH/22 Guard against possible overflow in moan_check_errorcopy().
237 PH/23 POSIX allows open() to be a macro; guard against that.
239 PH/24 If the recipient of an error message contained an @ in the local part
240 (suitably quoted, of course), incorrect values were put in $domain and
241 $local_part during the evaluation of errors_copy.
247 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
248 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
251 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
252 issue a MAIL command.
254 PH/02 In an ACL statement such as
256 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
258 if a client was not listed at all, or was listed with a value other than
259 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
260 the condition was not true (as it should be), so access was not denied.
261 The bug was that the ! inversion was incorrectly passed on to the second
262 item. This has been fixed.
264 PH/03 Added additional dnslists conditions == and =& which are different from
265 = and & when the dns lookup returns more than one IP address.
267 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
268 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
270 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
271 FSYNC, which compiles an option called disable_fsync that allows for
272 bypassing fsync(). The documentation is heavily laced with warnings.
274 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
276 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
277 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
278 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
279 including adding "make clean"; (3) Added -fPIC when compiling the test
280 dynamically loaded module, to get rid of a warning.
282 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
283 message fails, move_frozen_messages = true and ignore_bounce_errors_after
284 = 0s. The bug is otherwise harmless.
286 PH/07 There was a bug in the dovecot authenticator such that the value of
287 $auth1 could be overwritten, and so not correctly preserved, after a
288 successful authentication. This usually meant that the value preserved by
289 the server_setid option was incorrect.
291 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
293 PH/09 Installed PCRE release 7.0.
295 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
296 run for batched SMTP input. It is now run at the start of every message
297 in the batch. While fixing this I discovered that the process information
298 (output by running exiwhat) was not always getting set for -bs and -bS
299 input. This is fixed, and it now also says "batched" for BSMTP.
301 PH/11 Added control=no_pipelining.
303 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
304 patch, slightly modified), and move the expansion of helo_data till after
305 the connection is made in the smtp transport (so it can use these
308 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
310 PH/14 Added log_selector = +pid.
312 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
314 PH/16 Add ${if forany and ${if forall.
316 PH/17 Added dsn_from option to vary the From: line in DSNs.
318 PH/18 Flush SMTP output before performing a callout, unless control =
319 no_callout_flush is set.
321 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
322 was true (the default) a successful delivery failed to delete the retry
323 item, thus causing premature timeout of the address. The bug is now
326 PH/20 Added hosts_avoid_pipelining to the smtp transport.
328 PH/21 Long custom messages for fakedefer and fakereject are now split up
329 into multiline reponses in the same way that messages for "deny" and
330 other ACL rejections are.
332 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
333 with slight modification.
335 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
336 draft, changing the syntax and factoring some duplicate code".
338 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
339 for deliveries of the second and subsequent messages over the same SMTP
342 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
343 ${reduce, with only minor "tidies".
345 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
347 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
348 expansion side effects.
350 PH/27 When a message times out after an over-quota error from an Exim-imposed
351 quota, the bounce message says "mailbox is full". This message was not
352 being given when it was a system quota that was exceeded. It now should
355 MH/03 Made $recipients available in local_scan(). local_scan() already has
356 better access to the recipient list through recipients_list[], but
357 $recipients can be useful in postmaster-provided expansion strings.
359 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
360 in the case of a MAIL command with additional options following the
361 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
362 were accidentally chopped off.
364 PH/29 SMTP synchronization checks are implemented when a command is read -
365 there is a check that no more input is waiting when there shouldn't be
366 any. However, for some commands, a delay in an ACL can mean that it is
367 some time before the response is written. In this time, more input might
368 arrive, invalidly. So now there are extra checks after an ACL has run for
369 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
370 pipelining has not been advertised.
372 PH/30 MH's patch to allow iscntrl() characters to be list separators.
374 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
375 returned in the SMTP response when smtp_return_error_details was false.
378 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
379 instead of the C I/O that was originally supplied, because problems were
382 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
383 Exim which did not show up earlier: it was assuming that a call to
384 SSL_CTX_set_info_callback() might give an error value. In fact, there is
385 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
386 was a macro that became an assignment, so it seemed to work. This has
387 changed to a proper function call with a void return, hence the compile
388 error. Exim's code has been fixed.
390 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
393 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
395 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
397 PH/37 If a message is not accepted after it has had an id assigned (e.g.
398 because it turns out to be too big or there is a timeout) there is no
399 "Completed" line in the log. When some messages of this type were
400 selected by exigrep, they were listed as "not completed". Others were
401 picked up by some special patterns. I have improved the selection
402 criteria to be more general.
404 PH/38 The host_find_failed option in the manualroute router can now be set
405 to "ignore", to completely ignore a host whose IP address cannot be
406 found. If all hosts are ignored, the behaviour is controlled by the new
407 host_all_ignored option.
409 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
410 homing or because of multiple MX records with /mx) generated more than
411 one IP address, and the following item turned out to be the local host,
412 all the secondary addresses of the first item were incorrectly removed
413 from the list, along with the local host and any following hosts (which
414 is what is supposed to happen).
416 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
417 whoever called Exim into the -H file. In the case of the daemon it was
418 behaving confusingly. When first started, it used values for whoever
419 started the daemon, but after a SIGHUP it used the Exim user (because it
420 calls itself on a restart). I have changed the code so that it now always
423 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
424 message are rejected with the same error (e.g. no authentication or bad
425 sender address), and a DATA command is nevertheless sent (as can happen
426 with PIPELINING or a stupid MUA), the error message that was given to the
427 RCPT commands is included in the rejection of the DATA command. This is
428 intended to be helpful for MUAs that show only the final error to their
431 PH/42 Another patch from the Sieve maintainer.
433 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
434 Eximstats - Fixed some broken HTML links and added missing column headers
436 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
437 columns for Rejects, Temp Rejects, Ham, and Spam rows.
439 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
441 PH/43 Yet another patch from the Sieve maintainer.
443 PH/44 I found a way to check for a TCP/IP connection going away before sending
444 the response to the final '.' that terminates a message, but only in the
445 case where the client has not sent further data following the '.'
446 (unfortunately, this is allowed). However, in many cases there won't be
447 any further data because there won't be any more messages to send. A call
448 to select() can be used: if it shows that the input is "ready", there is
449 either input waiting, or the socket has been closed. An attempt to read
450 the next input character can distinguish the two cases. Previously, Exim
451 would have sent an OK response which the client would never have see.
452 This could lead to message repetition. This fix should cure that, at
453 least in a lot of common cases.
455 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
456 advertised in response to EHLO.
462 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
463 fixed by 4.65/MH/01 (is this a record?) are fixed:
465 (i) An empty string was always treated as zero by the numeric comparison
466 operators. This behaviour has been restored.
468 (ii) It is documented that the numeric comparison operators always treat
469 their arguments as decimal numbers. This was broken in that numbers
470 starting with 0 were being interpreted as octal.
472 While fixing these problems I realized that there was another issue that
473 hadn't been noticed. Values of message_size_limit (both the global option
474 and the transport option) were treated as octal if they started with 0.
475 The documentation was vague. These values are now always treated as
476 decimal, and I will make that clear in the documentation.
482 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
483 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
486 MH/01 Don't check that the operands of numeric comparison operators are
487 integers when their expansion is in "skipping" mode (fixes bug
488 introduced by 4.64-PH/07).
490 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
491 child addresses, Exim now panics and dies. Previously, because the count
492 is held in a short int, deliveries were likely to be lost. As such a
493 large number of recipients for a single message is ridiculous
494 (performance will be very, very poor), I have chosen to impose a limit
495 rather than extend the field.
501 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
502 leftover -K file (the existence of which was triggered by #402).
503 While we were at it, introduced process PID as part of the -K
504 filename. This should rule out race conditions when creating
507 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
508 processing considerably. Previous code took too long for large mails,
509 triggering a timeout which in turn triggers #401.
511 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
512 in the DK code in transports.c. sendfile() is not really portable,
513 hence the _LINUX specificness.
515 TF/01 In the add_headers option to the mail command in an Exim filter,
516 there was a bug that Exim would claim a syntax error in any
517 header after the first one which had an odd number of characters
520 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
521 callout verification, Exim cached a "reject" for the entire domain. This
522 is correct for most verifications, but it is not correct for a recipient
523 verification with use_sender or use_postmaster set, because in that case
524 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
525 case of MAIL FROM:<> rejection from other early rejections (e.g.
526 rejection of HELO). When verifying a recipient using a non-null MAIL
527 address, the cache is ignored if it shows MAIL FROM:<> rejection.
528 Whatever the result of the callout, the value of the domain cache is
529 left unchanged (for any other kind of callout, getting as far as trying
530 RCPT means that the domain itself is ok).
532 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
535 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
536 manifest itself as EPIPE rather than ECONNECT. When tidying away a
537 session, the daemon ignores ECONNECT errors and logs others; it now
538 ignores EPIPE as well.
540 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
541 (quoted-printable decoding).
543 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
544 later the small subsequent patch to fix an introduced bug.
546 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
548 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
550 PH/08 An error is now given if message_size_limit is specified negative.
552 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
553 to be given (somewhat) arbitrary names.
555 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
558 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
559 miscellaneous code fixes
561 PH/10 Added the log_reject_target ACL modifier to specify where to log
564 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
565 hostname. This is wrong, because it relates to the incoming message (and
566 probably the interface on which it is arriving) and not to the outgoing
567 callout (which could be using a different interface). This has been
568 changed to use the value of the helo_data option from the smtp transport
569 instead - this is what is used when a message is actually being sent. If
570 there is no remote transport (possible with a router that sets up host
571 addresses), $smtp_active_hostname is used.
573 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
574 tweaks were necessary in order to get it to work (see also 21 below):
575 (a) The code assumed that strncpy() returns a negative number on buffer
576 overflow, which isn't the case. Replaced with Exim's string_format()
578 (b) There were several signed/unsigned issues. I just did the minimum
579 hacking in of casts. There is scope for a larger refactoring.
580 (c) The code used strcasecmp() which is not a standard C function.
581 Replaced with Exim's strcmpic() function.
582 (d) The code set only $1; it now sets $auth1 as well.
583 (e) A simple test gave the error "authentication client didn't specify
584 service in request". It would seem that Dovecot has changed its
585 interface. Fortunately there's a specification; I followed it and
586 changed what the client sends and it appears to be working now.
588 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
591 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
592 address A is aliased to B and C, where B exists and C does not. Without
593 -v the output is "A verified" because verification stops after a
594 successful redirection if more than one address is generated. However,
595 with -v the child addresses are also verified. Exim was outputting "A
596 failed to verify" and then showing the successful verification for C,
597 with its parentage. It now outputs "B failed to verify", showing B's
598 parentage before showing the successful verification of C.
600 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
601 look up a TXT record in a specific list after matching in a combined
604 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
605 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
606 they consult the DNS. I had assumed they would set it the way they
607 wanted; and indeed my experiments on Linux seem to show that in some
608 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
609 To be on the safe side, however, I have now made the interface to
610 host_find_byname() similar to host_find_bydns(), with an argument
611 containing the DNS resolver options. The host_find_byname() function now
612 sets these options at its start, just as host_find_bydns() does. The smtp
613 transport options dns_qualify_single and dns_search_parents are passed to
614 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
615 of host_find_byname() use the default settings of RES_DEFNAMES
616 (qualify_single) but not RES_DNSRCH (search_parents).
618 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
619 spool_read_header() do less string testing, by means of a preliminary
620 switch on the second character of optional "-foo" lines. (This is
621 overdue, caused by the large number of possibilities that now exist.
622 Originally there were few.) While I was there, I also converted the
623 str(n)cmp tests so they don't re-test the leading "-" and the first
624 character, in the hope this might squeeze out yet more improvement.
626 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
627 flag allowing group syntax was set by the header_syntax check but not
628 turned off, possible causing trouble later; (2) The flag was not being
629 set at all for the header_verify test, causing "group"-style headers to
630 be rejected. I have now set it in this case, and also caused header_
631 verify to ignore an empty address taken from a group. While doing this, I
632 came across some other cases where the code for allowing group syntax
633 while scanning a header line wasn't quite right (mostly, not resetting
634 the flag correctly in the right place). These bugs could have caused
635 trouble for malformed header lines. I hope it is now all correct.
637 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
638 with the "reply" argument non-NULL. The code, however (which originally
639 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
640 but it didn't always do it. This confused somebody who was copying the
641 code for some other use. I have removed all the tests.
643 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
644 feature that was used to support insecure browsers during the U.S. crypto
645 embargo. It requires special client support, and Exim is probably the
646 only MTA that supported it -- and would never use it because real RSA is
647 always available. This code has been removed, because it had the bad
648 effect of slowing Exim down by computing (never used) parameters for the
649 RSA_EXPORT functionality.
651 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
652 authenticator to fail if there's a tab character in the incoming data
653 (there should never be unless someone is messing about, as it's supposed
654 to be base64-encoded). Also added, on Timo's advice, the "secured" option
655 if the connection is using TLS or if the remote IP is the same as the
656 local IP, and the "valid-client-cert option" if a client certificate has
659 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
660 authenticators. This can be used for authorization after authentication
661 succeeds. (In the case of plaintext, it servers for both authentication
664 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
665 if any retry times were supplied.
667 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
668 connection, where there is no client IP address to check. In this
669 situation, the verify now always succeeds.
671 PH/25 Applied John Jetmore's -Mset patch.
673 PH/26 Added -bem to be like -Mset, but loading a message from a file.
675 PH/27 In a string expansion for a processed (not raw) header when multiple
676 headers of the same name were present, leading whitespace was being
677 removed from all of them, but trailing whitespace was being removed only
678 from the last one. Now trailing whitespace is removed from each header
679 before concatenation. Completely empty headers in a concatenation (as
682 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
683 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
685 PH/29 [Removed. This was a change that I later backed out, and forgot to
686 correct the ChangeLog entry (that I had efficiently created) before
687 committing the later change.]
689 PH/30 Exim was sometimes attempting to deliver messages that had suffered
690 address errors (4xx response to RCPT) over the same connection as other
691 messages routed to the same hosts. Such deliveries are always "forced",
692 so retry times are not inspected. This resulted in far too many retries
693 for the affected addresses. The effect occurred only when there were more
694 hosts than the hosts_max_try setting in the smtp transport when it had
695 the 4xx errors. Those hosts that it had tried were not added to the list
696 of hosts for which the message was waiting, so if all were tried, there
697 was no problem. Two fixes have been applied:
699 (i) If there are any address or message errors in an SMTP delivery, none
700 of the hosts (tried or untried) are now added to the list of hosts
701 for which the message is waiting, so the message should not be a
702 candidate for sending over the same connection that was used for a
703 successful delivery of some other message. This seems entirely
704 reasonable: after all the message is NOT "waiting for some host".
705 This is so "obvious" that I'm not sure why it wasn't done
706 previously. Hope I haven't missed anything, but it can't do any
707 harm, as the worst effect is to miss an optimization.
709 (ii) If, despite (i), such a delivery is accidentally attempted, the
710 routing retry time is respected, so at least it doesn't keep
711 hammering the server.
713 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
714 in ${readsocket because some servers need this prod.
716 PH/32 Added some extra debug output when updating a wait-xxx database.
718 PH/33 The hint "could be header name not terminated by colon", which has been
719 given for certain expansion errors for a long time, was not being given
720 for the ${if def:h_colon_omitted{... case.
722 PH/34 The spec says: "With one important exception, whenever a domain list is
723 being scanned, $domain contains the subject domain." There was at least
724 one case where this was not true.
726 PH/35 The error "getsockname() failed: connection reset by peer" was being
727 written to the panic log as well as the main log, but it isn't really
728 panic-worthy as it just means the connection died rather early on. I have
729 removed the panic log writing for the ECONNRESET error when getsockname()
732 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
733 runs only) independently of the message's sender address. This meant
734 that, if the 4xx error was in fact related to the sender, a different
735 message to the same recipient with a different sender could confuse
736 things. In particualar, this can happen when sending to a greylisting
737 server, but other circumstances could also provoke similar problems.
738 I have changed the default so that the retry time for these errors is now
739 based a combination of the sender and recipient addresses. This change
740 can be overridden by setting address_retry_include_sender=false in the
743 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
744 remote server are returned as part of bounce messages. This was not
745 happening for LMTP over a pipe (the lmtp transport), but now it is the
746 same for both kinds of LMTP.
748 PH/38 Despite being documented as not happening, Exim was rewriting addresses
749 in header lines that were in fact CNAMEs. This is no longer the case.
751 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
752 and queue runs started by the daemon processed all messages. This has
753 been fixed so that -R and -S can now usefully be given with -q<time>.
755 PH/40 Import PCRE release 6.7 (fixes some bugs).
757 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
759 PH/42 Give an error if -q is specified more than once.
761 PH/43 Renamed the variables $interface_address and $interface_port as
762 $received_ip_address and $received_port, to make it clear that these
763 values apply to message reception, and not to the outgoing interface when
764 a message is delivered. (The old names remain recognized, of course.)
766 PH/44 There was no timeout on the connect() call when using a Unix domain
767 socket in the ${readsocket expansion. There now is.
769 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
770 be meaningful with "accept".
772 SC/01 Eximstats V1.43
773 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
775 SC/02 Eximstats V1.44
776 Use a glob alias rather than an array ref in the generated
777 parser. This improves both readability and performance.
779 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
780 Collect SpamAssassin and rejection statistics.
781 Don't display local sender or destination tables unless
782 there is data to show.
783 Added average volumes into the top table text output.
785 SC/04 Eximstats V1.46
786 Collect data on the number of addresses (recipients)
787 as well as the number of messages.
789 SC/05 Eximstats V1.47
790 Added 'Message too big' to the list of mail rejection
791 reasons (thanks to Marco Gaiarin).
793 SC/06 Eximstats V1.48
794 Mainlog lines which have GMT offsets and are too short to
795 have a flag are now skipped.
797 SC/07 Eximstats V1.49 (Alain Williams)
798 Added the -emptyok flag.
800 SC/08 Eximstats V1.50
801 Fixes for obtaining the IP address from reject messages.
803 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
804 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
805 whitesspace changes from 4.64-PH/27
807 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
810 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
811 are found, allow negative numbers in numeric criteria)
813 JJ/06 exipick.20061117.2, added new $message_body_missing variable
815 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
816 to match changes made in 4.64-PH/43
818 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
820 PH/47 Put in an explicit test for a DNS lookup of an address record where the
821 "domain" is actually an IP address, and force a failure. This locks out
822 those revolvers/nameservers that support "A-for-A" lookups, in
823 contravention of the specifications.
825 PH/48 When a host name was looked up from an IP address, and the subsequent
826 forward lookup of the name timed out, the host name was left in
827 $sender_host_name, contrary to the specification.
829 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
830 restricted to single-key lookups, Exim was not diagnosing an error if
831 * or *@ was used with a query-style lookup.
833 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
835 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
836 long ago, but noone interested enough thought of it. Let's just say that
837 the "1.1" means that there are some new functions that weren't there at
838 some point in the past.
840 PH/51 Error processing for expansion failure of helo_data from an smtp
841 transport during callout processing was broken.
843 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
844 tested/used via the -bh/-bhc/-bs options.
846 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
847 bug, fixed in subsequent PCRE releases).
849 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
850 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
852 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
858 SC/01 Use a glob alias rather than an array ref in eximstats generated
859 parser. This improves both readability and performance.
861 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
862 Don't display local sender or destination tables in eximstats unless
863 there is data to show.
864 Added average volumes into the eximstats top table text output.
866 SC/03 Collect data on the number of addresses (recipients) as well
867 as the number of messages in eximstats.
869 TF/01 Correct an error in the documentation for the redirect router. Exim
870 does (usually) call initgroups() when daemonizing.
872 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
873 with consistent privilege compared to when running as a daemon.
875 TF/03 Note in the spec that $authenticated_id is not set for local
876 submissions from trusted users.
878 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
879 Thanks to Dean Brooks <dean@iglou.com> for the patch.
881 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
882 by adding some example configuration directives to the default
883 configuration file. A little bit of work is required to uncomment the
884 directives and define how usernames and passwords are checked, but
885 there is now a framework to start from.
887 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
888 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
889 without this. I don't know how relevant this is to other LDAP libraries.
891 PH/02 Add the verb name to the "unknown ACL verb" error.
893 PH/03 Magnus Holmgren's patch for filter_prepend_home.
895 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
897 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
898 directory not expanded when it should be if an expanded home directory
899 was set for the address (which is overridden by the transport).
901 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
904 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
905 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
906 because it is too late at that time, and has no effect.
908 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
909 security issue with \' (bugzilla #107). I could not use the
910 PQescapeStringConn() function, because it needs a PGconn value as one of
913 PH/08 When testing addresses using -bt, indicate those final addresses that
914 are duplicates that would not cause an additional delivery. At least one
915 person was confused, thinking that -bt output corresponded to deliveries.
916 (Suppressing duplicates isn't a good idea as you lose the information
917 about possibly different redirections that led to the duplicates.)
919 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
920 systems where poll() doesn't work, in particular OS X.
922 PH/10 Added more information to debugging output for retry time not reached.
924 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
925 operations in malware.c.
927 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
930 PH/13 If write_rejectlog was set false when logging was sent to syslog with
931 syslog_duplication set false, log lines that would normally be written
932 both the the main log and to the reject log were not written to syslog at
935 PH/14 In the default configuration, change the use of "message" in ACL warn
936 statements to "add_header".
938 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
939 not followed by a command (e.g. "seen endif").
941 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
942 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
945 PH/17 Added extra conditions to the default value of delay_warning_condition
949 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
950 { match{$h_precedence:}{(?i)bulk|list|junk} } \
951 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
954 The Auto-Submitted: and various List- headers are standardised, whereas I
955 don't think Precedence: ever was.
957 PH/18 Refactored debugging code in route_finduser() to show more information,
958 in particular, the error code if getpwnam() issues one.
960 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
961 This is apparently needed in addition to the PH/07 change above to avoid
962 any possible encoding problems.
964 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
965 but not after initializing Perl.
967 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
968 output them only if debugging. By default they are written stderr,
969 apparently, which is not desirable.
971 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
974 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
977 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
979 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
980 authenticated or an ident call has been made. Suppress the default
981 values for $authenticated_id and $authenticated_sender (but permit -oMai
982 and -oMas) when testing with -bh.
984 PH/24 Re-jigged the order of the tests in the default configuration so that the
985 tests for valid domains and recipients precede the DNS black list and CSA
986 tests, on the grounds that those ones are more expensive.
988 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
989 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
990 This bug exists in every version of Exim that I still have, right back to
993 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
994 However, an attempt to turn on case-sensitivity in a regex key by
995 including (?-i) didn't work because the subject string was already
996 lowercased, and the effects were non-intuitive. It turns out that a
997 one-line patch can be used to allow (?-i) to work as expected.
1003 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1004 other effects) broke the use of negated acl sub-conditions.
1006 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1009 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1010 "Deny" causes Exim to reject the incoming connection with a 554 error.
1011 Unfortunately, if there is a major crisis, such as a disk failure,
1012 tcp-wrappers gives "deny", whereas what one would like would be some
1013 kind of temporary error. A kludge has been added to help with this.
1014 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1015 554 error is used if errno is still zero or contains ENOENT (which occurs
1016 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1019 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1021 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1022 errors. Otherwise a message that provokes a temporary error (when other
1023 messages do not) can cause a whole host to time out.
1025 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1026 addresses were routed directly to files or pipes from a redirect router.
1027 File deliveries just didn't batch; pipe deliveries might have suffered
1030 PH/06 A failure to get a lock for a hints database would erroneously always say
1031 "Failed to get write lock", even when it was really a read lock.
1033 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1034 of 0600. This has been changed to use the value of the lockfile_mode
1035 option (which defaults to 0600).
1037 PH/08 Applied small patch from the Sieve maintainer.
1039 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1040 folder from quota calculations, a direct delivery into this folder messed
1041 up the contents of the maildirsize file. This was because the regex was
1042 used only to exclude .Trash (or whatever) when the size of the mailbox
1043 was calculated. There was no check that a delivery was happening into an
1044 excluded directory. This bug has been fixed by ignoring all quota
1045 processing for deliveries into excluded directories.
1047 PH/10 Added the maildirfolder_create_regex option to appendfile.
1053 PH/01 The code for finding all the local interface addresses on a FreeBSD
1054 system running IPv6 was broken. This may well have applied to all BSD
1055 systems, as well as to others that have similar system calls. The broken
1056 code found IPv4 interfaces correctly, but gave incorrect values for the
1057 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1058 that it would not match correctly against @[] and not recognize the IPv6
1061 PH/02 The ipliteral router was not recognizing addresses of the form user@
1062 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1064 PH/03 Added disable_ipv6.
1066 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1067 decoded form, because it is most often used to construct To: headers
1068 lines in autoreplies, and the decoded form may well be syntactically
1069 invalid. However, $reply_address has leading white space removed, and all
1070 newlines turned into spaces so that the autoreply transport does not
1073 PH/05 If group was specified without a user on a router, and no group or user
1074 was specified on a transport, the group from the router was ignored.
1076 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1077 for visible compile-time settings that can be used to change these
1078 numbers, for those that want even more. Backwards compatibility with old
1079 spool files has been maintained. However, going back to a previous Exim
1080 release will lost any variables that are in spool files.
1082 PH/07 Two small changes when running in the test harness: increase delay when
1083 passing a TCP/IP connection to a new process, in case the original
1084 process has to generate a bounce, and remove special handling of
1085 127.0.0.2 (sic), which is no longer necessary.
1087 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1088 be the same on different OS.
1090 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1093 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1094 whether --show-vars was specified or not
1096 JJ/02 exipick: Added support for new ACL variable spool format introduced
1099 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1100 syntactically invalid From: or Reply-to: line, and a filter used this to
1101 generate an autoreply, and therefore failed to obtain an address for the
1102 autoreply, Exim could try to deliver to a non-existent relative file
1103 name, causing unrelated and misleading errors. What now happens is that
1104 it logs this as a hard delivery error, but does not attempt to create a
1107 PH/11 The exinext utility has a -C option for testing purposes, but although
1108 the given file was scanned by exinext itself; it wasn't being passed on
1109 when Exim was called.
1111 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1112 an end-of-file indication when reading a command response.
1114 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1115 compiled. In many other places in Exim, IPv6 addresses are always
1116 recognized, so I have changed this. It also means that IPv4 domain
1117 literals of the form [IPV4:n.n.n.n] are now always recognized.
1119 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1120 used if the router is not running as root, for example, when verifying at
1121 ACL time, or when using -bh. The debugging output from this situation was
1122 non-existent - all you got was a failure to exec. I have made two
1125 (a) Failures to set uid/gid, the current directory, or a process leader
1126 in a subprocess such as that created by queryprogram now generate
1127 suitable debugging ouput when -d is set.
1129 (b) The queryprogram router detects when it is not running as root,
1130 outputs suitable debugging information if -d is set, and then runs
1131 the subprocess without attempting to change uid/gid.
1133 PH/15 Minor change to Makefile for building test_host (undocumented testing
1136 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1137 additional section of a DNS packet that returns MX or SRV records.
1138 Instead, it always explicitly searches for A/AAAA records. This avoids
1139 major problems that occur when a DNS server includes only records of one
1140 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1141 fixed another bug: if SRV records were looked up and the corresponding
1142 address records were *not* found in the additional section, the port
1143 values from the SRV records were lost.
1145 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1146 using the correct key (the original address) when searching the retry
1147 rules in order to find which one to use for generating the retry hint.
1149 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1150 adding the default one. Similarly, if it contains a Reply-To: header, the
1151 errors_reply_to option, if set, is not used.
1153 PH/19 When calculating a retry time, Exim used to measure the "time since
1154 failure" by looking at the "first failed" field in the retry record. Now
1155 it does not use this if it is later than than the arrival time of the
1156 message. Instead it uses the arrival time. This makes for better
1157 behaviour in cases where some deliveries succeed, thus re-setting the
1158 "first failed" field. An example is a quota failure for a huge message
1159 when small messages continue to be delivered. Without this change, the
1160 "time since failure" will always be short, possible causing more frequent
1161 delivery attempts for the huge message than are intended.
1162 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1164 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1165 $1, $2, $3) because the numerical variables can be reset during some
1166 expansion items (e.g. "match"), thereby losing the authentication data.
1168 PH/21 Make -bV show the size of off_t variables so that the test suite can
1169 decide whether to run tests for quotas > 2G.
1171 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1172 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1173 filecount value is greater than 2G or if a quota value is greater than 2G
1174 on a system where the size of off_t is not greater than 4, a panic error
1177 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1178 never match. The debug and -bh output now contains an explicit error
1179 message indicating a malformed IPv4 address or mask.
1181 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1182 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1183 PH/23 above applies.
1185 PH/25 Do not write to syslog when running in the test harness. The only
1186 occasion when this arises is a failure to open the main or panic logs
1187 (for which there is an explicit test).
1189 PH/26 Added the /no_tell option to "control=freeze".
1191 PH/27 If a host name lookup failed very early in a connection, for example, if
1192 the IP address matched host_lookup and the reverse lookup yielded a name
1193 that did not have a forward lookup, an error message of the form "no IP
1194 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1195 could be logged. Now it outputs the IP address instead of "NULL".
1197 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1198 allows the sender and the authenticated sender to be set when
1199 submitting a message from within Exim. Since child_open_exim() is
1200 documented for local_scan(), the new function should be too.
1202 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1203 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1204 results in an empty string is now treated as unset.
1206 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1208 PH/31 Added sender_verify_fail logging option.
1210 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1211 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1212 tidied the source and removed it altogether.
1214 PH/33 When a queue run was abandoned because the load average was too high, a
1215 log line was always written; now it is written only if the queue_run log
1216 selector is set. In addition, the log line for abandonment now contains
1217 information about the queue run such as the pid. This is always present
1218 in "start" and "stop" lines but was omitted from the "abandon" line.
1220 PH/34 Omit spaces between a header name and the colon in the error message that
1221 is given when verify = headers_syntax fails (if there are lots of them,
1222 the message gets confusing).
1224 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1225 names, as there are now some PTR records that contain slashes. This check
1226 is only to protect against broken name servers that fall over on strange
1227 characters, so the fact that it applies to all lookups doesn't matter.
1229 PH/36 Now that the new test suite is complete, we can remove some of the
1230 special code in Exim that was needed for the old test suite. For example,
1231 sorting DNS records because real resolvers return them in an arbitrary
1232 order. The new test suite's fake resolver always returns records in the
1235 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1236 bounces) except when queue_only is set, to avoid logging races between
1237 the different processes.
1239 PH/38 Panic-die if .include specifies a non-absolute path.
1241 PH/39 A tweak to the "H" retry rule from its user.
1243 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1244 a label. They prevented compilation on older perls.
1246 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1247 a warning to be raised on newish perls.
1249 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1250 on queue. Changes to match documented behaviour of showing count of
1251 messages matching specified criteria.
1253 PH/40 Changed the default ident timeout from 30s to 5s.
1255 PH/41 Added support for the use of login_cap features, on those BSD systems
1256 that have them, for controlling the resources used by pipe deliveries.
1258 PH/42 The content-scanning code uses fopen() to create files in which to put
1259 message data. Previously it was not paying any attention to the mode of
1260 the files. Exim runs with umask(0) because the rest of the code creates
1261 files with open(), and sets the required mode explicitly. Thus, these
1262 files were ending up world-writeable. This was not a big issue, because,
1263 being within the spool directory, they were not world-accessible. I have
1264 created a function called modefopen, which takes an additional mode
1265 argument. It sets umask(777), creates the file, chmods it to the required
1266 mode, then resets the umask. All the relevant calls to fopen() in the
1267 content scanning code have been changed to use this function.
1269 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1270 to 24 hours. This avoids potential overflow problems when processing G
1271 and H retry rules. I suspect nobody ever tinkers with this value.
1273 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1275 PH/45 When the plaintext authenticator is running as a client, the server's
1276 challenges are checked to ensure they are valid base64 strings. By
1277 default, the authentication attempt is cancelled if an invalid string is
1278 received. Setting client_ignore_invalid_base64 true ignores these errors.
1279 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1280 they are received. Thus, the responses can be made to depend on the
1281 challenges. If an invalid string is ignored, an empty string is placed in
1284 PH/46 Messages that are created by the autoreply transport now contains a
1285 References: header, in accordance with RFCs 2822 and 3834.
1287 PH/47 Added authenticated_sender_force to the smtp transport.
1289 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1291 PH/49 Installed latest patch from the Sieve maintainer.
1293 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1294 was also set, the appendfile transport was unnecessarily scanning a
1295 directory of message files (e.g. for maildir delivery) to find the count
1296 of files (along with the size), even though it did not need this
1297 information. It now does the scan only if it needs to find either the
1298 size of the count of files.
1300 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1302 PH/52 Two bugs concerned with error handling when the smtp transport is
1305 (i) Exim was not creating retry information for temporary errors given
1306 for individual recipients after the DATA command when the smtp transport
1307 was used in LMTP mode. This meant that they could be retried too
1308 frequently, and not timed out correctly.
1310 (ii) Exim was setting the flag that allows error details to be returned
1311 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1312 recipients that were returned after the DATA command.
1314 PH/53 This is related to PH/52, but is more general: for any failing address,
1315 when detailed error information was permitted to be returned to the
1316 sender, but the error was temporary, then after the final timeout, only
1317 "retry timeout exceeded" was returned. Now it returns the full error as
1318 well as "retry timeout exceeded".
1320 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1321 do this, and (what is worse) MTAs that accept it.
1323 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1324 will now be deprecated.
1326 PH/56 New os.c-cygwin from the Cygwin maintainer.
1328 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1329 formats (previously only available in exim formats via -bpr, -bpru,
1330 and -bpra. Now also available in native and exiqgrep formats)
1332 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1333 with very large, slow to parse queues
1335 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1337 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1339 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1340 responses to authentication challenges, though it was showing the
1341 challenges; (ii) I've removed the CR characters from the debug output for
1344 PH/58 Allow for the insertion of a newline as well as a space when a string
1345 is turned into more than one encoded-word during RFC 2047 encoding. The
1346 Sieve code now uses this.
1348 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1349 data_4xx, lost_connection, tls_required.
1351 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1352 message was being sent as an SMTP response.
1354 PH/61 Add -l and -k options to exicyclog.
1356 PH/62 When verifying, if an address was redirected to one new address, so that
1357 verification continued, and the new address failed or deferred after
1358 having set something in $address_data, the value of $address_data was not
1359 passed back to the ACL. This was different to the case when no
1360 redirection occurred. The value is now passed back in both cases.
1362 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1363 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1364 use login_cap.h, so on its own it isn't the distinguishing feature. The
1365 new name refers directly to the setclassresources() function.
1367 PH/65 Added configuration files for NetBSD3.
1369 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1371 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1372 is preferred over IPv4.
1374 PH/68 The bounce_return_message and bounce_return_body options were not being
1375 honoured for bounces generated during the reception of non-SMTP messages.
1376 In particular, this applied to messages rejected by the ACL. This bug has
1377 been fixed. However, if bounce_return_message is true and bounce_return_
1378 body is false, the headers that are returned for a non-SMTP message
1379 include only those that have been read before the error was detected.
1380 (In the case of an ACL rejection, they have all been read.)
1382 PH/69 The HTML version of the specification is now built in a directory called
1383 spec_html instead of spec.html, because the latter looks like a path with
1384 a MIME-type, and this confuses some software.
1386 PH/70 Catch two compiler warnings in sieve.c.
1388 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1389 function verify_get_ident() calls ip_connect() to connect a socket, but
1390 if the "connect()" function timed out, ip_connect() used to close the
1391 socket. However, verify_get_ident() also closes the socket later, and in
1392 between Exim writes to the log, which may get opened at this point. When
1393 the socket was closed in ip_connect(), the log could get the same file
1394 descriptor number as the socket. This naturally causes chaos. The fix is
1395 not to close the socket in ip_connect(); the socket should be closed by
1396 the function that creates it. There was only one place in the code where
1397 this was missing, in the iplookup router, which I don't think anybody now
1398 uses, but I've fixed it anyway.
1400 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1401 well as to direct DNS lookups. Otherwise the handling of names in host
1402 lists is inconsistent and therefore confusing.
1408 PH/01 Two changes to the default runtime configuration:
1410 (1) Move the checks for relay_from_hosts and authenticated clients from
1411 after to before the (commented out) DNS black list checks.
1413 (2) Add control=submission to the relay_from_hosts and authenticated
1414 clients checks, on the grounds that messages accepted by these
1415 statements are most likely to be submissions.
1417 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1419 (1) Generate an error if the third argument for the ${prvs expansion is
1422 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1425 (3) Reset the variables that are obtained from the first argument of
1426 ${prvscheck and used in the second argument before leaving the code,
1427 because their memory is reclaimed, so using them afterwards may do
1430 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1431 one (it's much easier than Tom thought :-).
1433 (5) Because of (4), we can now allow for the use of $prvscheck_result
1434 inside the third argument.
1436 PH/03 For some reason, the default setting of PATH when running a command from
1437 a pipe transport was just "/usr/bin". I have changed it to
1440 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1441 anything to be listed in the output from -bV.
1443 PH/05 When a filter generated an autoreply, the entire To: header line was
1444 quoted in the delivery log line, like this:
1446 => >A.N.Other <ano@some.domain> <original@ddress> ...
1448 This has been changed so that it extracts the operative address. There
1449 may be more than one such address. If so, they are comma-separated, like
1452 => >ano@some.domain,ona@other.domain <original@ddress> ...
1454 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1455 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1456 not being looked up in the rDNS to get a host name, Exim was showing the
1457 IP address twice in Received: lines, even though the IP addresses were
1458 identical. For example:
1460 Received: from [1.2.3.4] (helo=[1.2.3.4])
1462 However, if the real host name was known, it was omitting the HELO data
1463 if it matched the actual IP address. This has been tidied up so that it
1464 doesn't show the same IP address twice.
1466 PH/07 When both +timestamp and +memory debugging was on, the value given by
1467 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1468 called by the debug printing, thereby overwriting the timestamp buffer.
1469 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1471 PH/08 When the original message was included in an autoreply transport, it
1472 always said "this is a copy of the message, including all the headers",
1473 even if body_only or headers_only was set. It now gives an appropriate
1476 PH/09 Applied a patch from the Sieve maintainer which:
1478 o fixes some comments
1479 o adds the (disabled) notify extension core
1480 o adds some debug output for the result of if/elsif tests
1481 o points to the current vacation draft in the documentation
1482 and documents the missing references header update
1486 o fixes a bug in processing the envelope test (when testing
1487 multiple envelope elements, the last element determinted the
1490 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1491 Electronic Mail") by including:
1493 Auto-submitted: auto-generated
1495 in the messages that it generates (bounce messages and others, such as
1496 warnings). In the case of bounce messages for non-SMTP mesages, there was
1497 also a typo: it was using "Auto_submitted" (underscore instead of
1498 hyphen). Since every message generated by Exim is necessarily in response
1499 to another message, thes have all been changed to:
1501 Auto-Submitted: auto-replied
1503 in accordance with these statements in the RFC:
1505 The auto-replied keyword:
1507 - SHOULD be used on messages sent in direct response to another
1508 message by an automatic process,
1510 - MUST NOT be used on manually-generated messages,
1512 - MAY be used on Delivery Status Notifications (DSNs) and Message
1513 Disposition Notifications (MDNs),
1515 - MUST NOT be used on messages generated by automatic or periodic
1516 processes, except for messages which are automatic responses to
1519 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1520 to the default Received: header definition.
1522 PH/12 Added log selector acl_warn_skipped (default on).
1524 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1525 variables because (a) they are in the wrong storage pool and (b) even if
1526 they were copied, it wouldn't work properly because of the caching.
1528 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1529 checking when decoding. Apparently there are clients that generate
1530 overlong encoded strings. Why am I not surprised?
1532 PH/15 If the first argument of "${if match_address" was not empty, but did not
1533 contain an "@" character, Exim crashed. Now it writes a panic log message
1534 and treats the condition as false.
1536 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1538 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1539 extension "envelope-auth". The code is finished and in agreement with
1540 other implementations, but there is no documentation so far and in fact,
1541 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1542 not changing the active code.
1544 Print executed "if" and "elsif" statements when debugging is used. This
1545 helps a great deal to understand what a filter does.
1547 Document more things not specified clearly in RFC3028. I had all this
1548 sorted out, when out of a sudden new issues came to my mind. Oops."
1550 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1553 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1554 canonical form (as documented). However, after a host name lookup from
1555 the IP address, check_host() was doing a simple string comparison with
1556 addresses acquired from the DNS when checking that the found name did
1557 have the original IP as one of its addresses. Since any found IPv6
1558 addresses are likely to be in abbreviated form, the comparison could
1559 fail. Luckily, there already exists a function for doing the comparison
1560 by converting both addresses to binary, so now that is used instead of
1561 the text comparison.
1563 PH/20 There was another similar case to PH/19, when a complete host name was
1564 given in a host list; looking up its IP address could give an abbreviated
1565 form, whereas the current host's name might or might not be abbreviated.
1566 The same fix has been applied.
1572 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1573 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1576 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1577 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1579 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1581 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1582 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1583 Also, for "Auto-Submitted", treat anything other than "no" as
1584 non-personal, in accordance with RFC 3834. (Previously it treated
1585 anything starting "auto-" as non-personal.)
1587 TF/01 The control=submission/name=... option had a problem with syntax
1588 errors if the name included a slash character. The /name= option
1589 now slurps the rest of the string, so it can include any characters
1590 but it must come last in the list of options (after /sender_retain
1593 PH/05 Some modifications to the interface to the fake nameserver for the new
1601 TK/01 Added the "success_on_redirect" address verification option. See
1602 NewStuff for rationale and an example.
1604 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1606 PH/02 Patch to exigrep to allow it to work on syslog lines.
1608 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1609 fread() to skip over the body file's header line, because in Cygwin the
1610 header line is locked and is inaccessible.
1612 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1613 co-exist for some time) to make it clear that it is the Exim ID that is
1614 referenced, not the Message-ID: header line.
1616 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1617 string_format() function, because snprintf() does not exist on all
1620 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1621 ${stat: expansion item.
1623 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1624 protocol synchronization error", to keep the pedants happy.
1626 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1627 well as for IRIX systems, when gcc is being used. See the host.c source
1630 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1632 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1635 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1636 transport and to the smtp transport in LMTP mode.
1638 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1640 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1641 run a filter in a subprocess. This could lead to confusion in subsequent
1642 lookups in the parent process. There should also be a search_tidyup() at
1643 the end of the subprocess.
1645 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1646 only if the host matched helo_try_verify_hosts, which caused the
1647 verification to occur when the EHLO/HELO command was issued. The ACL just
1648 tested the remembered result. Now, if a previous verification attempt has
1649 not happened, "verify = helo" does it there and then.
1651 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1653 TK/03 Fix log output including CR from clamd.
1655 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1656 exist provoked a memory error which could cause a segfault.
1658 PH/15 Installed PCRE 6.2
1660 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1662 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1663 of the problem. Specifically, suggested +O2 rather than +O1 for the
1666 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1668 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1669 record, the port was not correctly logged when the outgoing_port log
1670 selector was set (it logged the transort's default port).
1672 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1673 fallback_hosts, and "hosts" in the smtp transport.
1675 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1676 host errors such as "Connection refused".
1678 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1679 authentication with radiusclient 0.4.9:
1681 - Error returned from rc_read_config was caught wrongly
1682 - Username/password not passed on to radius server due to wrong length.
1684 The presumption is that some radiusclient API changes for 4.51/PH/17
1685 were not taken care of correctly. The code is still untested by me (my
1686 Linux distribution still has 0.3.2 of radiusclient), but it was
1687 contributed by a Radius user.
1689 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1690 expanding the "port" option of the smtp transport.
1692 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1693 while reading a MIME header. Thanks to Tom Hughes for a patch.
1695 PH/24 Include config.h inside local_scan.h so that configuration settings are
1698 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1699 that in an ACL for RCPT (for example), you can examine exactly what was
1702 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1703 commands, but it was not correctly comparing the address with the actual
1704 client host address. Thus, it would show the EHLO address in Received:
1705 header lines when this was not necessary.
1707 PH/27 Added the % operator to ${eval:}.
1709 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1710 it should be ignoring failures (because with -C, for example, it has lost
1711 privilege). It wasn't ignoring creation failures other than "already
1714 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1717 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1718 because an input line was too long, either on its own, or by virtue of
1719 too many continuations, the temporary file was not being removed, and the
1720 return code was incorrect.
1722 PH/31 Missing "BOOL" in function definition in filtertest.c.
1724 PH/32 Applied Sieve patches from the maintainer.
1726 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1728 PH/33 Added "verify = not_blind".
1730 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1731 Local/Makefile (with some defaults set). These are used in built scripts
1732 such as exicyclog, but they have never been used in the exim_install
1733 script (though there are many overriding facilities there). I have
1734 arranged that the exim_install script now takes note of these two
1737 PH/35 Installed configuration files for Dragonfly.
1739 PH/36 When a locally submitted message by a trusted user did not contain a
1740 From: header, and the sender address was obtained from -f or from an SMTP
1741 MAIL command, and the trusted user did not use -F to supply a sender
1742 name, $originator_name was incorrectly used when constructing a From:
1743 header. Furthermore, $originator_name was used for submission mode
1744 messages from external hosts without From: headers in a similar way,
1745 which is clearly wrong.
1747 PH/37 Added control=suppress_local_fixups.
1749 PH/38 When log_selector = +received_sender was set, and the addition of the
1750 sender made the log line's construction buffer exactly full, or one byte
1751 less than full, an overflow happened when the terminating "\n" was
1754 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1755 when the result of a list match is failure because a DNS lookup failed.
1757 PH/40 RM_COMMAND is now used in the building process.
1759 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1760 the "build-* directories that it finds.
1762 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1763 address in a domain literal was a prefix of an interface address.
1765 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1766 when verifying a sender address, unless rewrite_headers is false.
1768 PH/44 Wrote a long comment about why errors_to addresses are verified as
1769 recipients, not senders.
1771 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1772 the ratelimit ACL was added.
1774 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1776 PH/46 Added extra information about PostgreSQL errors to the error string.
1778 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1779 avoiding the need to install special zones in a real server. This is
1780 backwards compatible; if it can't find the fake resolver, it drops back.
1781 Thus, both old and new test suites can be run.
1783 TF/02 Added util/ratelimit.pl
1785 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1786 clock is set back in time.
1788 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1789 Candler <B.Candler@pobox.com>.
1791 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1792 OK for addresses that are the result of redirections.
1794 PH/48 A number of further additions for the benefit of the new test suite,
1795 including a fake gethostbyname() that interfaces to the fake DNS resolver
1798 TF/06 The fix for widen_domains has also been applied to qualify_single and
1799 search_parents which are the other dnslookup options that can cause
1802 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1805 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1807 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1808 service" if it encountered an MX record with an empty target hostname.
1809 The message is now "an MX or SRV record indicated no SMTP service".
1811 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1812 if the DNS of the sending site is misconfigured. This is quite a
1813 common situation. This change restores the behaviour of treating a
1814 helo verification defer as a failure.
1816 PH/51 If self=fail was set on a router, the bounce message did not include the
1817 actual error message.
1823 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1825 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1826 command itself ended in error, the underlying message about the transport
1827 filter timeout was being overwritten with the pipe command error. Now the
1828 underlying error message should be appended to the second error message.
1830 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1832 PH/02 Reduce the amount of output that "make" produces by default. Full output
1833 can still be requested.
1835 PH/03 The warning log line about a condition test deferring for a "warn" verb
1836 was being output only once per connection, rather than after each
1837 occurrence (because it was using the same function as for successful
1838 "warn" verbs). This seems wrong, so I have changed it.
1840 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1841 it should not have, which might have caused a crash in the right
1842 circumstances, but probably never did.
1844 PH/04 Installed a modified version of Tony Finch's patch to make submission
1845 mode fix the return path as well as the Sender: header line, and to
1846 add a /name= option so that you can make the user's friendly name appear
1849 TF/03 Added the control = fakedefer ACL modifier.
1851 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1852 Mark Lowes for thorough testing.
1854 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1856 TK/03 Merged latest SRS patch from Miles Wilton.
1858 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1859 with the definition in sysexits.h (which is #included earlier).
1860 Fortunately, Exim does not actually use EX_OK. The code used to try to
1861 preserve the sysexits.h value, by assumimg that macro definitions were
1862 scanned for macro replacements. I have been disabused of this notion,
1863 so now the code just undefines EX_OK before #including unistd.h.
1865 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1866 in the smtp transport. When a block could not be written in a single
1867 write() function, the timeout was being re-applied to each part-write.
1868 This seems wrong - if the receiver was accepting one byte at a time it
1869 would take for ever. The timeout is now adjusted when this happens. It
1870 doesn't have to be particularly precise.
1872 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1873 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1875 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1876 without a domain if the check to <postmaster@domain> fails.
1878 SC/01 Eximstats: added -xls and the ability to specify output files
1879 (patch written by Frank Heydlauf).
1881 SC/02 Eximstats: use FileHandles for outputing results.
1883 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1885 SC/04 Eximstats: fixed display of large numbers with -nvr option
1887 SC/05 Eximstats: fixed merging of reports with empty tables.
1889 SC/06 Eximstats: added the -include_original_destination flag
1891 SC/07 Eximstats: removed tabs and trailing whitespace.
1893 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1895 TK/06 MBOX spool code: Add real "From " MBOX separator line
1896 so the .eml file is really in mbox format (even though
1897 most programs do not really care). Patch from Alex Miller.
1899 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1900 The latter is generated from $received_to and is only set if the
1901 message has one envelope recipient. SA can use these headers,
1902 obviously out-of-the-box. Patch from Alex Miller.
1904 PH/08 The ${def test on a variable was returning false if the variable's
1905 value was "0", contrary to what the specification has always said!
1906 The result should be true unless the variable is empty.
1908 PH/09 The syntax error of a character other than { following "${if
1909 def:variable_name" (after optional whitespace) was not being diagnosed.
1910 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1911 accidental colon was present, for example, could give incorrect results.
1913 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1914 result is used (not including appendfile, where other changes are about
1917 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1918 This involved changing a lot of size variables from int to off_t. It
1919 should work with maildirs and everything.
1921 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1922 spamd dying while we are connected to it.
1924 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1927 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1928 function for BDB 4.3.
1930 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1932 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1933 the sources for PCRE have been re-arranged and more files are now
1936 PH/15 The code I had for printing potentially long long variables in PH/11
1937 above was not the best (it lost precision). The length of off_t variables
1938 is now inspected at build time, and an appropriate printing format (%ld
1939 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1940 to be "long long int" or "long int". This is needed for the internal
1941 formatting function string_vformat().
1943 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1944 the configuration file to be ":syslog", then the script "guesses" where
1945 the logs files are, rather than using the compiled in default. In our
1946 case the guess is not the same as the compiled default, so the script
1947 suddenly stopped working when I started to use syslog. The patch checks
1948 to see if log_file_path is "". If so, it attempts to read it from exim
1949 with no configuration file to get the compiled in version, before it
1950 falls back to the previous guessing code."
1952 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1953 implementing BATV in an Exim configuration. See NewStuff for the gory
1956 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1957 Makefile that are specific to HP-UX.
1959 PH/18 If the "use_postmaster" option was set for a recipient callout together
1960 with the "random" option, the postmaster address was used as the MAIL
1961 FROM address for the random test, but not for the subsequent recipient
1962 test. It is now used for both.
1964 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1965 patch removes a few documentation additions to RFC 3028, because the
1966 latest draft now contains them. It adds the new en;ascii-case comparator
1967 and a new error check for 8bit text in MIME parts. Comparator and
1968 require names are now matched exactly. I enabled the subaddress
1969 extension, but it is not well tested yet (read: it works for me)."
1971 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1972 rework some of the code of TK/09 above to avoid the hardwired use of
1973 "%lld" and "long long". Replaced the call to snprintf() with a call to
1976 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1977 records point to non-existent hosts", "retry timeout exceeded", and
1978 "retry time not reached for any host after a long failure period".
1980 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1981 experimental DomainKeys support:
1983 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1984 (2) On an error such as an illegally used "control", the wrong name for
1985 the control was given.
1987 These problems did NOT occur unless DomainKeys support was compiled.
1989 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1991 PH/24 Added ${if match_ip condition.
1993 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1994 databases so that it will be absolutely obvious if a crash occurs in the
1995 DB library. This is a regular occurrence (often caused by mis-matched
1998 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1999 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2000 picked up on a user's system that detects such things. There doesn't seem
2001 to be a gcc warning option for this - only an attribute that has to be
2002 put on the function's prototype. It seems that in Fedora Core 4 they have
2003 set this on a number of new functions. No doubt there will be more in due
2006 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2007 specify a transport. However, if an address that was verified by such a
2008 router was the subject of a callout, Exim crashed because it tried to
2009 read the rcpt_include_affixes from the non-existent transport. Now it
2010 just assumes that the setting of that option is false. This bug was
2011 introduced by 4.51/PH/31.
2013 PH/28 Changed -d+all to exclude +memory, because that information is very
2014 rarely of interest, but it makes the output a lot bigger. People tend to
2015 do -d+all out of habit.
2017 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2018 code in os-type was giving problems when libc.so lives in lib64, like on
2021 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2022 aren't the modern standard, and it seems that some systems' include files
2023 don't always have them. Exim was already checking for some of the newer
2024 ones like T_AAAA, and defining it itself. I've added checks for all the
2025 record types that Exim uses.
2027 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2028 not automatically generating a new one, as it is supposed to. This
2029 prevented TLS from working. If the file did exist, but contained invalid
2030 data, a new version was generated, as expected. It was only the case of a
2031 non-existent file that was broken.
2033 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2034 with a change in libdomainkeys > 0.64.
2036 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2037 from DNS. If the selector record carries the flag, it now has
2038 precedence over the domain-wide flag.
2040 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2042 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2043 the use of a port name that isn't defined in /etc/services) occurred, the
2044 message was deferred as in a normal delivery, and thus remained on the
2045 spool, instead of being failed because of the mua_wrapper setting. This
2046 is now fixed, and I tidied up some of the mua_wrapper messages at the
2049 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2050 the messages in a hash of arrays rather than using individual hashes.
2051 This is a bit cleaner and results in dramatic memory savings, albeit
2052 at a slight CPU cost.
2054 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2055 as requested by Marc Sherman.
2057 SC/10 Eximstats: added histograms for user specified patterns as requested
2060 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2062 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2063 fopen() in the content-scanning modules that did not already have it.
2069 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2070 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2072 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2074 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2076 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2077 to test Sieve filters that use "vacation".
2079 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2080 that changes the way the GnuTLS parameters are stored in the cache file.
2081 The new format can be generated externally. For backward compatibility,
2082 if the data in the cache doesn't make sense, Exim assumes it has read an
2083 old-format file, and it generates new data and writes a new file. This
2084 means that you can't go back to an older release without removing the
2087 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2088 work if there are any delivery delays because "one_time" forces the
2089 parent to be marked "delivered", so its unseen clone is never tried
2090 again. For this reason, Exim now forbids the simultaneous setting of
2093 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2094 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2095 entry if you want to know the details. The fix, however, neglected to
2096 consider the case when local delivery batching is involved. The test for
2097 "previously delivered" was not happening when checking to see if an
2098 address could be batched with a previous (undelivered) one; under
2099 certain circumstances this could lead to multiple deliveries to the same
2102 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2103 in its include files, and this causes problems building Exim.
2105 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2106 header_syntax) but Exim was just ignoring anything given after a slash.
2107 In particular, this caused confusion with an attempt to use "verify =
2108 reverse_host_lookup/defer_ok". An error is now given when options are
2109 supplied for verify items that do not have them. (Maybe reverse_host_
2110 lookup should have a defer_ok option, but that's a different point.)
2112 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2113 defined by RFC 821) to 2048, because there were problems with some AUTH
2114 commands, and RFC 1869 says the size should be increased for extended
2115 SMTP commands that take arguments.
2117 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2120 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2121 "unknown" error; now it says that the functionality isn't in the binary.
2123 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2124 an address' error message when a string expansion fails (syntax or
2125 whatever). Otherwise the password may appear in the log. Following change
2126 PH/42 below, there is no longer a chance of it appearing in a bounce
2129 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2131 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2132 of its addresses was ever tried. (Bugzilla bug #2.)
2134 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2135 the result incorrectly in the debug output. (It correctly added a newline
2136 to what was transported.)
2138 TF/01 Added $received_time.
2140 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2141 commented out examples of how to interface to a virus scanner and to
2142 SpamAssassin. Also added commented examples of av_scanner and
2143 spamd_address settings.
2145 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2146 and controls are allowed in which ACLs. There were a couple of minor
2147 errors. Some of the entries in the conditions table (which is a table of
2148 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2149 negation of where the condition IS allowed.
2151 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2153 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2154 header file does not have a version number, so I've had to invent a new
2155 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2156 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2157 radiusclient), but it was contributed by a Radius user.
2159 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2160 files or directories, for OpenSSL.
2162 PH/19 When an Exim process that is running as root has to create an Exim log
2163 file, it does so in a subprocess that runs as exim:exim so as to get the
2164 ownership right at creation (otherwise, other Exim processes might see
2165 the file with the wrong ownership). There was no test for failure of this
2166 fork() call, which would lead to the process getting stuck as it waited
2167 for a non-existent subprocess. Forks do occasionally fail when resources
2168 run out. I reviewed all the other calls to fork(); they all seem to check
2171 PH/20 When checking for unexpected SMTP input at connect time (before writing
2172 the banner), Exim was not dealing correctly with a non-positive return
2173 from the read() function. If the client had disconnected by this time,
2174 the result was a log entry for a synchronization error with an empty
2175 string after "input=" when read() returned zero. If read() returned -1
2176 (an event I could not check), uninitialized data bytes were printed.
2177 There were reports of junk text (parts of files, etc) appearing after
2180 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2182 PH/22 Added support for macro redefinition, and (re)definition in between
2183 driver and ACL definitions.
2185 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2186 forgetting to use the resulting value; it was using the unexpanded value.
2188 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2189 hadn't been configured. The fix is from Juergen Kreileder, who
2190 understands it better than I do:
2192 "Here's what I see happening with three configured cyrus_sasl
2193 authenticators configured (plain, login, cram-md5):
2195 On startup auth_cyrus_sasl_init() gets called for each of these.
2196 This means three calls to sasl_listmech() without a specified mech_list.
2197 => SASL tests which mechs of all available mechs actually work
2198 => three warnings about OTP not working
2199 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2201 With the patch, sasl_listmech() also gets called three times. But now
2202 SASL's mech_list option is set to the server_mech specified in the the
2203 authenticator. Or in other words, the answer from sasl_listmech()
2204 gets limited to just the mech you're testing for (which is different
2206 => the return list contains just 'plain' or 'login', 'cram-md5' or
2207 nothing depending on the value of ob->server_mech.
2209 I've just tested the patch: Authentication still works fine,
2210 unavailable mechs specified in the exim configuration are still
2211 caught, and the auth.log warnings about OTP are gone."
2213 PH/25 When debugging is enabled, the contents of the command line are added
2214 to the debugging output, even when log_selector=+arguments is not
2217 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2218 answer is "GNU", and only if the return is "GNU/something" is the answer
2221 PH/27 $acl_verify_message is now set immediately after the failure of a
2222 verification in an ACL, and so is available in subsequent modifiers. In
2223 particular, the message can be preserved by coding like this:
2225 warn !verify = sender
2226 set acl_m0 = $acl_verify_message
2228 Previously, $acl_verify_message was set only while expanding "message"
2229 and "log_message" when a very denied access.
2231 PH/28 Modified OS/os.c-Linux with
2233 -#ifndef OS_LOAD_AVERAGE
2234 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2236 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2237 nomenclature these days.)
2239 PH/29 Installed patch from the Sieve maintainer that adds the options
2240 sieve_useraddress and sieve_subaddress to the redirect router.
2242 PH/30 In these circumstances:
2243 . Two addresses routed to the same list of hosts;
2244 . First host does not offer TLS;
2245 . First host accepts first address;
2246 . First host gives temporary error to second address;
2247 . Second host offers TLS and a TLS session is established;
2248 . Second host accepts second address.
2249 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2250 and peerdn, if requested) that were in fact used only for the second
2253 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2254 attention to any local part prefix or suffix that was matched by the
2255 router that accepted the address. It now behaves in the same way as it
2256 does for delivery: the affixes are removed from the local part unless
2257 rcpt_include_affixes is set on the transport.
2259 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2260 timeout during the DATA phase of an incoming message.
2262 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2263 applied a patch sanctioned by the Sieve maintainer.
2265 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2266 the uid or gid is negative. A case of a negative gid caused this to be
2267 noticed. The fix allows for either to be negative.
2269 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2270 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2273 PH/36 The change PH/12 above was broken. Fixed it.
2275 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2276 the grounds that routing the same address twice would always produce the
2277 same answer. This might have been true once, but it is certainly no
2278 longer true now. Routing a child address may depend on the previous
2279 routing that produced that child. Some complicated redirection strategies
2280 went wrong when messages had multiple recipients, and made Exim's
2281 behaviour dependent on the order in which the addresses were given.
2283 I have moved the duplicate checking until after the routing is complete.
2284 Exim scans the addresses that are assigned to local and remote
2285 transports, and removes any duplicates. This means that more work will be
2286 done, as duplicates will always all be routed, but duplicates are
2287 presumably rare, so I don't expect this is of any significance.
2289 For deliveries to pipes, files, and autoreplies, the duplicate checking
2290 still happens during the routing process, since they are not going to be
2293 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2294 It corrects a timeout issue with spamd. This is Ian's comment: "The
2295 background is that sometimes spamd either never reads data from a
2296 connection it has accepted, or it never writes response data. The exiscan
2297 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2298 blindly assumes that writes won't block so it may never time out."
2300 PH/39 Allow G after quota size as well as K and M.
2302 PH/40 The value set for $authenticated_id in an authenticator may not contain
2303 binary zeroes or newlines because the value is written to log lines and
2304 to spool files. There was no check on this. Now the value is run through
2305 the string_printing() function so that such characters are converted to
2306 printable escape sequences.
2308 PH/41 $message_linecount is a new variable that contains the total number of
2309 lines in the message. Compare $body_linecount, which is the count for the
2312 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2313 bounce and delay warning messages, except in certain special cases, which
2316 (a) An SMTP error message from a remote host;
2317 (b) A message specified in a :fail: redirection;
2318 (c) A message specified in a "fail" command in a system filter;
2319 (d) A message specified in a FAIL return from the queryprogram router;
2320 (e) A message specified by the cannot_route_message router option.
2322 In these cases only, Exim does include the error details in bounce and
2323 warning messages. There are also a few cases where bland messages such
2324 as "unrouteable address" or "local delivery error" are given.
2326 PH/43 $value is now also set for the "else" part of a ${run expansion.
2328 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2329 being worked on, but at least Exim now implements the latest version to
2332 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2333 process to complete was treated as a delivery failure, a timeout while
2334 writing the message to the pipe was logged, but erroneously treated as a
2335 successful delivery. Such timeouts include transport filter timeouts. For
2336 consistency with the overall process timeout, these timeouts are now
2337 treated as errors, giving rise to delivery failures by default. However,
2338 there is now a new Boolean option for the pipe transport called
2339 timeout_defer, which, if set TRUE, converts the failures into defers for
2340 both kinds of timeout. A transport filter timeout is now identified in
2343 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2344 systems where "make" and "gmake" are different, calling "gmake" at top
2345 level broke things. I've arranged for the value of $(MAKE) to be passed
2346 from the Makefile to this script so that it can call the same version of
2350 A note about Exim versions 4.44 and 4.50
2351 ----------------------------------------
2353 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2354 changes of various kinds. As a consequence, a big documentation update was
2355 needed. This delayed the release for rather longer than seemed good, especially
2356 in the light of a couple of (minor) security issues. Therefore, the changes
2357 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2358 release. So 4.44 and 4.50 are in effect two different branches that both start
2361 I have left the 4.50 change log unchanged; it contains all the changes since
2362 4.43. The change log for 4.44 is below; many of its items are identical to
2363 those for 4.50. This seems to be the most sensible way to preserve the
2364 historical information.
2370 1. Minor wording change to the doc/README.SIEVE file.
2372 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2373 computation of the current number of files was incorrect.
2375 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2376 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2379 4. Give more explanation in the error message when the command for a transport
2380 filter fails to execute.
2382 5. There are several places where Exim runs a non-Exim command in a
2383 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2384 was being done only for the command run by the queryprogram router. It is
2385 now done for all such subprocesses. The other cases are: ${run, transport
2386 filters, and the commands run by the lmtp and pipe transports.
2388 6. Added CONFIGURE_GROUP build-time option.
2390 7. Some older OS have a limit of 256 on the maximum number of file
2391 descriptors. Exim was using setrlimit() to set 1000 as a large value
2392 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2393 systems. I've change it so that if it can't get 1000, it tries for 256.
2395 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2396 was an oversight, and furthermore, ever since the addition of extra
2397 controls (e.g. 4.43/32), the checks on when to allow different forms of
2398 "control" were broken. There should now be diagnostics for all cases when a
2399 control that does not make sense is encountered.
2401 9. Added the /retain_sender option to "control=submission".
2403 10. $recipients is now available in the predata ACL (oversight).
2405 11. Tidy the search cache before the fork to do a delivery from a message
2406 received from the command line. Otherwise the child will trigger a lookup
2407 failure and thereby defer the delivery if it tries to use (for example) a
2408 cached ldap connection that the parent has called unbind on.
2410 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2411 of $address_data from the recipient verification was clobbered by the
2412 sender verification.
2414 13. The value of address_data from a sender verification is now available in
2415 $sender_address_data in subsequent conditions in the ACL statement.
2417 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2419 15. Added a new option "connect=<time>" to callout options, to set a different
2422 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2423 was its contents. (It was OK if the option was not defined at all.)
2425 17. A "Completed" log line is now written for messages that are removed from
2426 the spool by the -Mrm option.
2428 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2429 information about exactly what failed.
2431 19. Added -dd to debug only the daemon process.
2433 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2434 handles timeouts, both on the server side and network timeouts. Renamed the
2435 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2437 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2438 It is now set to "smtps".
2440 22. $host_address is now set to the target address during the checking of
2441 ignore_target_hosts.
2443 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2444 being passed; this would have caused $sender_host_name to have been used if
2445 matching the list had actually called for a host name (not very likely,
2446 since this list is usually IP addresses). A host name is now passed as
2449 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2450 code that specifies a non-restarting handler (typically sigaction() in
2451 modern systems) in an attempt to fix a rare and obscure crash bug.
2453 25. Narrowed the window for a race in the daemon that could cause it to ignore
2454 SIGCHLD signals. This is not a major problem, because they are used only to
2455 wake it up if nothing else does.
2457 26. A malformed maildirsize file could cause Exim to calculate negative values
2458 for the mailbox size or file count. Odd effects could occur as a result.
2459 The maildirsize information is now recalculated if the size or filecount
2462 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2463 support for a long time. Removed HAVE_SYS_VFS_H.
2465 28. Installed the latest version of exipick from John Jetmore.
2467 29. In an address list, if the pattern was not a regular expression, an empty
2468 subject address (from a bounce message) matched only if the pattern was an
2469 empty string. Non-empty patterns were not even tested. This was the wrong
2470 because it is perfectly reasonable to use an empty address as part of a
2471 database query. An empty address is now tested by patterns that are
2472 lookups. However, all the other forms of pattern expect the subject to
2473 contain a local part and a domain, and therefore, for them, an empty
2474 address still always fails if the pattern is not itself empty.
2476 30. Exim went into a mad DNS loop when attempting to do a callout where the
2477 host was specified on an smtp transport, and looking it up yielded more
2478 than one IP address.
2480 31. Re-factored the code for checking spool and log partition space into a
2481 function that finds that data and another that does the check. The former
2482 is then used to implement four new variables: $spool_space, $log_space,
2483 $spool_inodes, and $log_inodes.
2485 32. The RFC2047 encoding function was originally intended for short strings
2486 such as real names; it was not keeping to the 75-character limit for
2487 encoded words that the RFC imposes. It now respects the limit, and
2488 generates multiple encoded words if necessary. To be on the safe side, I
2489 have increased the buffer size for the ${rfc2047: expansion operator from
2492 33. It is now permitted to omit both strings after an "if" condition; if the
2493 condition is true, the result is "true". As before, when the second string
2494 is omitted, a false condition yields an empty string. This makes it less
2495 cumbersome to write custom ACL and router conditions.
2497 34. Failure to deliver a bounce message always caused it to be frozen, even if
2498 there was an errors_to setting on the router. The errors_to setting is now
2501 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2502 canonical form (fully expanded) before being placed in
2503 $sender_host_address.
2505 36. The table in the code that translates DNS record types into text (T_A to
2506 "A" for instance) was missing entries for NS and CNAME. It is just possible
2507 that this could have caused confusion if both these types were looked up
2508 for the same domain, because the text type is used as part of Exim's
2509 per-process caching. But the chance of anyone hitting this buglet seems
2512 37. The dnsdb lookup has been extended in a number of ways.
2514 (1) There is a new type, "zns", which walks up the domain tree until it
2515 finds some nameserver records. It should be used with care.
2517 (2) There is a new type, "mxh", which is like "mx" except that it returns
2518 just the host names, not the priorities.
2520 (3) It is now possible to give a list of domains (or IP addresses) to be
2521 looked up. The behaviour when one of the lookups defers can be
2522 controlled by a keyword.
2524 (4) It is now possible to specify the separator character for use when
2525 multiple records are returned.
2527 38. The dnslists ACL condition has been extended: it is now possible to supply
2528 a list of IP addresses and/or domains to be looked up in a particular DNS
2531 39. Added log_selector=+queue_time_overall.
2533 40. When running the queue in the test harness, wait just a tad after forking a
2534 delivery process, to get repeatability of debugging output.
2536 41. Include certificate and key file names in error message when GnuTLS fails
2537 to set them up, because the GnuTLS error message doesn't include the name
2538 of the failing file when there is a problem reading it.
2540 42. Allow both -bf and -bF in the same test run.
2542 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2544 44. The "Exiscan patch" is now merged into the mainline Exim source.
2546 45. Sometimes the final signoff response after QUIT could fail to get
2547 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2548 < 0 before doing a fflush(). This bug looks as though it goes back to the
2549 introduction of TLS in release 3.20, but "sometimes" must have been rare
2550 because the tests only now provoked it.
2552 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2553 (this can affect the format of dates).
2555 47. exim_tidydb, when checking for the continued existence of a message for
2556 which it has found a message-specific retry record, was not finding
2557 messages that were in split spool directories. Consequently, it was
2558 deleting retry records that should have stayed in existence.
2560 48. Steve fixed some bugs in eximstats.
2562 49. The SPA authentication driver was not abandoning authentication and moving
2563 on to the next authenticator when an expansion was forced to fail,
2564 contradicting the general specification for all authenticators. Instead it
2565 was generating a temporary error. It now behaves as specified.
2567 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2568 (the order specifies the preference for clients). The order is now AES256,
2569 AES128, 3DES, ARCFOUR128.
2571 51. Small patch to Sieve code - explicitly set From: when generating an
2574 52. Exim crashed if a remote delivery caused a very long error message to be
2575 recorded - for instance if somebody sent an entire SpamAssassin report back
2576 as a large number of 550 error lines. This bug was coincidentally fixed by
2577 increasing the size of one of Exim's internal buffers (big_buffer) that
2578 happened as part of the Exiscan merge. However, to be on the safe side, I
2579 have made the code more robust (and fixed the comments that describe what
2582 53. Now that there can be additional text after "Completed" in log lines (if
2583 the queue_time_overall log selector is set), a one-byte patch to exigrep
2584 was needed to allow it to recognize "Completed" as not the last thing in
2587 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2588 patch that reportedly fixes this has been added. I am not expert enough to
2589 create a test for it. This is what the patch creator wrote:
2591 "I found a little strange behaviour of ldap code when working with
2592 Windows 2003 AD Domain, where users was placed in more than one
2593 Organization Units. When I tried to give exim partial DN, the exit code
2594 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2595 But simultaneously result of request was absolutely normal ldap result,
2596 so I produce this patch..."
2598 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2599 REFERENCE, so I have modified the code to exclude the patch when that macro
2602 55. Some experimental protocols are using DNS PTR records for new purposes. The
2603 keys for these records are domain names, not reversed IP addresses. The
2604 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2605 leaves it alone. Component reversal etc. now happens only for IP addresses.
2608 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2610 57. Double the size of the debug message buffer (to 2048) so that more of very
2611 long debug lines gets shown.
2613 58. The exicyclog utility now does better if the number of log files to keep
2614 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2616 59. Two changes related to the smtp_active_hostname option:
2618 (1) $smtp_active_hostname is now available as a variable.
2619 (2) The default for smtp_banner uses $smtp_active_hostname instead
2620 of $primary_hostname.
2622 60. The host_aton() function is supposed to be passed a string that is known
2623 to be a valid IP address. However, in the case of IPv6 addresses, it was
2624 not checking this. This is a hostage to fortune. Exim now panics and dies
2625 if the condition is not met. A case was found where this could be provoked
2626 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2627 components; fortuitously, this particular loophole had already been fixed
2628 by change 4.50/55 above.
2630 If there are any other similar loopholes, the new check in host_aton()
2631 itself should stop them being exploited. The report I received stated that
2632 data on the command line could provoke the exploit when Exim was running as
2633 exim, but did not say which command line option was involved. All I could
2634 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2635 running as the user.
2638 61. There was a buffer overflow vulnerability in the SPA authentication code
2639 (which came originally from the Samba project). I have added a test to the
2640 spa_base64_to_bits() function which I hope fixes it.
2643 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2644 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2646 63. The daemon start-up calls getloadavg() while still root for those OS that
2647 need the first call to be done as root, but it missed one case: when
2648 deliver_queue_load_max is set with deliver_drop_privilege. This is
2649 necessary for the benefit of the queue runner, because there is no re-exec
2650 when deliver_drop_privilege is set.
2652 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2653 This has been fixed.
2655 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2656 was in use, was not putting the data itself into the right store pool;
2657 consequently, it could be overwritten for a subsequent message in the same
2658 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2661 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2663 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2664 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2665 the function were treating the return as a boolean value, which happened to
2666 work because 0=false and not-0=true, but is not correct code.
2668 68. The host_aton() function was not handling scoped IPv6 addresses (those
2669 with, for example, "%eth0" on the end) correctly.
2671 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2672 negated items (that is, ~something) in unsigned ints. Some compilers
2673 apparently mutter when there is no cast.
2675 70. If an address verification called from an ACL failed, and did not produce a
2676 user-specific message (i.e. there was only a "system" message), nothing was
2677 put in $acl_verify_message. In this situation, it now puts the system
2680 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2681 session; change 4.31/43 added the unwanted input to the log line - except
2682 that it did not do this in the start of session case. It now does.
2684 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2685 This is wrong and can cause the other end to generate a synchronization
2686 error if it is another Exim or anything else that does the synchronization
2687 check. A QUIT command is no longer sent after a timeout.
2689 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2690 during host lookups.
2692 74. The defer_ok option of callout verification was not working if it was used
2693 when verifying addresses in header lines, that is, for this case:
2695 verify = header_sender/callout=defer_ok
2697 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2698 those file descriptors could be used for SMTP connections. If anything
2699 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2700 could be sent to the SMTP client, causing chaos. The daemon now opens
2701 stdin, stdout, and stderr to /dev/null when it puts itself into the
2704 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2705 log by default. The user can override this with suitable Perl magic.
2707 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2708 the log message when discard triggers, was not working for the DATA ACL or
2709 for the non-SMTP ACL.
2711 78. Error message wording change in sieve.c.
2713 79. If smtp_accept_max_per_host was set, the number of connections could be
2714 restricted to fewer than expected, because the daemon was trying to set up
2715 a new connection before checking whether the processes handling previous
2716 connections had finished. The check for completed processes is now done
2717 earlier. On busy systems, this bug wouldn't be noticed because something
2718 else would have woken the daemon, and it would have reaped the completed
2721 80. If a message was submitted locally by a user whose login name contained one
2722 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2723 It caused a spool format error. I have fixed the spool reading code. A
2724 related problem was that the "from" clause in the Received: line became
2725 illegal because of the space(s). It is now covered by ${quote_local_part.
2727 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2730 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2732 83. Patch from Sieve maintainer to fix unterminated string problem in
2733 "vacation" handling.
2735 84. Some minor changes to the Linux configuration files to help with other
2736 OS variants using glibc.
2738 85. One more patch for Sieve to update vacation handling to latest spec.
2741 ----------------------------------------------------
2742 See the note above about the 4.44 and 4.50 releases.
2743 ----------------------------------------------------
2749 1. Change 4.43/35 introduced a bug that caused file counts to be
2750 incorrectly computed when quota_filecount was set in an appendfile
2753 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2754 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2757 3. Give more explanation in the error message when the command for a transport
2758 filter fails to execute.
2760 4. There are several places where Exim runs a non-Exim command in a
2761 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2762 was being done only for the command run by the queryprogram router. It is
2763 now done for all such subprocesses. The other cases are: ${run, transport
2764 filters, and the commands run by the lmtp and pipe transports.
2766 5. Some older OS have a limit of 256 on the maximum number of file
2767 descriptors. Exim was using setrlimit() to set 1000 as a large value
2768 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2769 systems. I've change it so that if it can't get 1000, it tries for 256.
2771 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2772 was an oversight, and furthermore, ever since the addition of extra
2773 controls (e.g. 4.43/32), the checks on when to allow different forms of
2774 "control" were broken. There should now be diagnostics for all cases when a
2775 control that does not make sense is encountered.
2777 7. $recipients is now available in the predata ACL (oversight).
2779 8. Tidy the search cache before the fork to do a delivery from a message
2780 received from the command line. Otherwise the child will trigger a lookup
2781 failure and thereby defer the delivery if it tries to use (for example) a
2782 cached ldap connection that the parent has called unbind on.
2784 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2785 of $address_data from the recipient verification was clobbered by the
2786 sender verification.
2788 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2789 was its contents. (It was OK if the option was not defined at all.)
2791 11. A "Completed" log line is now written for messages that are removed from
2792 the spool by the -Mrm option.
2794 12. $host_address is now set to the target address during the checking of
2795 ignore_target_hosts.
2797 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2798 being passed; this would have caused $sender_host_name to have been used if
2799 matching the list had actually called for a host name (not very likely,
2800 since this list is usually IP addresses). A host name is now passed as
2803 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2804 code that specifies a non-restarting handler (typically sigaction() in
2805 modern systems) in an attempt to fix a rare and obscure crash bug.
2807 15. Narrowed the window for a race in the daemon that could cause it to ignore
2808 SIGCHLD signals. This is not a major problem, because they are used only to
2809 wake it up if nothing else does.
2811 16. A malformed maildirsize file could cause Exim to calculate negative values
2812 for the mailbox size or file count. Odd effects could occur as a result.
2813 The maildirsize information is now recalculated if the size or filecount
2816 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2817 support for a long time. Removed HAVE_SYS_VFS_H.
2819 18. Updated exipick to current release from John Jetmore.
2821 19. Allow an empty sender to be matched against a lookup in an address list.
2822 Previously the only cases considered were a regular expression, or an
2825 20. Exim went into a mad DNS lookup loop when doing a callout where the
2826 host was specified on the transport, if the DNS lookup yielded more than
2829 21. The RFC2047 encoding function was originally intended for short strings
2830 such as real names; it was not keeping to the 75-character limit for
2831 encoded words that the RFC imposes. It now respects the limit, and
2832 generates multiple encoded words if necessary. To be on the safe side, I
2833 have increased the buffer size for the ${rfc2047: expansion operator from
2836 22. Failure to deliver a bounce message always caused it to be frozen, even if
2837 there was an errors_to setting on the router. The errors_to setting is now
2840 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2841 canonical form (fully expanded) before being placed in
2842 $sender_host_address.
2844 24. Updated eximstats to version 1.33
2846 25. Include certificate and key file names in error message when GnuTLS fails
2847 to set them up, because the GnuTLS error message doesn't include the name
2848 of the failing file when there is a problem reading it.
2850 26. Expand error message when OpenSSL has problems setting up cert/key files.
2853 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2854 (this can affect the format of dates).
2856 28. exim_tidydb, when checking for the continued existence of a message for
2857 which it has found a message-specific retry record, was not finding
2858 messages that were in split spool directories. Consequently, it was
2859 deleting retry records that should have stayed in existence.
2861 29. eximstats updated to version 1.35
2862 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2863 1.35 - bugfix such that pie charts by volume are generated correctly
2865 30. The SPA authentication driver was not abandoning authentication and moving
2866 on to the next authenticator when an expansion was forced to fail,
2867 contradicting the general specification for all authenticators. Instead it
2868 was generating a temporary error. It now behaves as specified.
2870 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2871 (the order specifies the preference for clients). The order is now AES256,
2872 AES128, 3DES, ARCFOUR128.
2874 31. Small patch to Sieve code - explicitly set From: when generating an
2877 32. Exim crashed if a remote delivery caused a very long error message to be
2878 recorded - for instance if somebody sent an entire SpamAssassin report back
2879 as a large number of 550 error lines. This bug was coincidentally fixed by
2880 increasing the size of one of Exim's internal buffers (big_buffer) that
2881 happened as part of the Exiscan merge. However, to be on the safe side, I
2882 have made the code more robust (and fixed the comments that describe what
2885 33. Some experimental protocols are using DNS PTR records for new purposes. The
2886 keys for these records are domain names, not reversed IP addresses. The
2887 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2888 leaves it alone. Component reversal etc. now happens only for IP addresses.
2891 34. The host_aton() function is supposed to be passed a string that is known
2892 to be a valid IP address. However, in the case of IPv6 addresses, it was
2893 not checking this. This is a hostage to fortune. Exim now panics and dies
2894 if the condition is not met. A case was found where this could be provoked
2895 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2896 components; fortuitously, this particular loophole had already been fixed
2897 by change 4.50/55 or 4.44/33 above.
2899 If there are any other similar loopholes, the new check in host_aton()
2900 itself should stop them being exploited. The report I received stated that
2901 data on the command line could provoke the exploit when Exim was running as
2902 exim, but did not say which command line option was involved. All I could
2903 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2904 running as the user.
2907 35. There was a buffer overflow vulnerability in the SPA authentication code
2908 (which came originally from the Samba project). I have added a test to the
2909 spa_base64_to_bits() function which I hope fixes it.
2912 36. The daemon start-up calls getloadavg() while still root for those OS that
2913 need the first call to be done as root, but it missed one case: when
2914 deliver_queue_load_max is set with deliver_drop_privilege. This is
2915 necessary for the benefit of the queue runner, because there is no re-exec
2916 when deliver_drop_privilege is set.
2918 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2919 was in use, was not putting the data itself into the right store pool;
2920 consequently, it could be overwritten for a subsequent message in the same
2921 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2924 38. Sometimes the final signoff response after QUIT could fail to get
2925 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2926 < 0 before doing a fflush(). This bug looks as though it goes back to the
2927 introduction of TLS in release 3.20, but "sometimes" must have been rare
2928 because the tests only now provoked it.
2934 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2935 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2936 Now it returns an integer. A number of calls were still expecting a T/F
2937 return. Fortuitously, in all cases, the tests worked in OK situations,
2938 which is the norm. However, things would have gone wrong on any write
2939 failures on the smtp file descriptor. This function is used when sending
2940 messages over SMTP and also when doing verify callouts.
2942 2. When Exim is called to do synchronous delivery of a locally submitted
2943 message (the -odf or -odi options), it no longer closes stderr before doing
2946 3. Implemented the mua_wrapper option.
2948 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2950 5. Implemented the functions header_remove(), header_testname(),
2951 header_add_at_position(), and receive_remove_recipient(), and exported them
2954 6. If an ACL "warn" statement specified the addition of headers, Exim already
2955 inserted X-ACL-Warn: at the start if there was no header name. However, it
2956 was not making this test for the second and subsequent header lines if
2957 there were newlines in the string. This meant that an invalid header could
2958 be inserted if Exim was badly configured.
2960 7. Allow an ACL "warn" statement to add header lines at the start or after all
2961 the Received: headers, as well as at the end.
2963 8. Added the rcpt_4xx retry error code.
2965 9. Added postmaster_mailfrom=xxx to callout verification option.
2967 10. Added mailfrom=xxxx to the callout verification option, for verify=
2970 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2971 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2973 12. Inserted some casts to stop certain compilers warning when using pointer
2974 differences as field lengths or precisions in printf-type calls (mostly
2975 affecting debugging statements).
2977 13. Added optional readline() support for -be (dynamically loaded).
2979 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2980 same clock tick as a message's arrival, so that its received time was the
2981 same as the "first fail" time on the retry record, and that message
2982 remained on the queue past the ultimate address timeout, every queue runner
2983 would try a delivery (because it was past the ultimate address timeout) but
2984 after another failure, the ultimate address timeout, which should have then
2985 bounced the address, did not kick in. This was a "< instead of <=" error;
2986 in most cases the first failure would have been in the next clock tick
2987 after the received time, and all would be well.
2989 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2990 being recognized when the domain list was tested by the match_domain
2991 condition in an expansion string.
2993 16. Added the ${str2b64: operator.
2995 17. Exim was always calling setrlimit() to set a large limit for the number of
2996 processes, without checking whether the existing limit was already
2997 adequate. (It did check for the limit on file descriptors.) Furthermore,
2998 errors from getrlimit() and setrlimit() were being ignored. Now they are
2999 logged to the main and panic logs, but Exim does carry on, to try to do its
3000 job under whatever limits there are.
3002 18. Imported PCRE 5.0.
3004 19. Trivial typo in log message " temporarily refused connection" (the leading
3007 20. If the log selector return_path_on_delivery was set and an address was
3008 redirected to /dev/null, the delivery process crashed because it assumed
3009 that a return path would always be set for a "successful" delivery. In this
3010 case, the whole delivery is bypassed as an optimization, and therefore no
3013 21. Internal re-arrangement: the function for sending a challenge and reading
3014 a response while authentication was assuming a zero-terminated challenge
3015 string. It's now changed to take a pointer and a length, to allow for
3016 binary data in such strings.
3018 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3020 23. Exim was not respecting finduser_retries when seeking the login of the
3021 uid under which it was called; it was always trying 10 times. (The default
3022 setting of finduser_retries is zero.) Also, it was sleeping after the final
3023 failure, which is pointless.
3025 24. Implemented tls_on_connect_ports.
3027 25. Implemented acl_smtp_predata.
3029 26. If the domain in control=submission is set empty, Exim assumes that the
3030 authenticated id is a complete email address when it generates From: or
3031 Sender: header lines.
3033 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3034 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3035 chown and chgrp in /bin and hostname in /usr/bin.
3037 28. Exim was keeping the "process log" file open after each use, just as it
3038 does for the main log. This opens the possibility of it remaining open for
3039 long periods when the USR1 signal hits a daemon. Occasional processlog
3040 errors were reported, that could have been caused by this. Anyway, it seems
3041 much more sensible not to leave this file open at all, so that is what now
3044 29. The long-running daemon process does not normally write to the log once it
3045 has entered its main loop, and it closes the log before doing so. This is
3046 so that log files can straightforwardly be renamed and moved. However,
3047 there are a couple of unusual error situations where the daemon does write
3048 log entries, and I had neglected to close the log afterwards.
3050 30. The text of an SMTP error response that was received during a remote
3051 delivery was being truncated at 512 bytes. This is too short for some of
3052 the long messages that one sometimes sees. I've increased the limit to
3055 31. It is now possible to make retry rules that apply only when a message has a
3056 specific sender, in particular, an empty sender.
3058 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3059 it possible to be selective about when SMTP synchronization is enforced.
3061 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3063 32. Implemented hosts_connection_nolog.
3065 33. Added an ACL for QUIT.
3067 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3070 35. Added mailbox_size and mailbox_filecount to appendfile.
3072 36. Added control = no_multiline_responses to ACLs.
3074 37. There was a bug in the logic of the code that waits for the clock to tick
3075 in the case where the clock went backwards by a substantial amount such
3076 that the microsecond fraction of "now" was more than the microsecond
3077 fraction of "then" (but the whole seconds number was less).
3079 38. Added support for the libradius Radius client library this is found on
3080 FreeBSD (previously only the radiusclient library was supported).
3086 1. When certain lookups returned multiple values in the form name=value, the
3087 quoting of the values was not always being done properly. Specifically:
3088 (a) If the value started with a double quote, but contained no whitespace,
3090 (b) If the value contained whitespace other than a space character (i.e.
3091 tabs or newlines or carriage returns) it was not quoted.
3092 This fix has been applied to the mysql and pgsql lookups by writing a
3093 separate quoting function and calling it from the lookup code. The fix
3094 should probably also be applied to nisplus, ibase and oracle lookups, but
3095 since I cannot test any of those, I have not disturbed their existing code.
3097 2. A hit in the callout cache for a specific address caused a log line with no
3098 reason for rejecting RCPT. Now it says "Previous (cached) callout
3099 verification failure".
3101 3. There was an off-by-one bug in the queryprogram router. An over-long
3102 return line was truncated at 256 instead of 255 characters, thereby
3103 overflowing its buffer with the terminating zero. As well as fixing this, I
3104 have increased the buffer size to 1024 (and made a note to document this).
3106 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3107 when Exim is waiting for an SMTP response from a remote server, Exim
3108 restarts its select() call on the socket, thereby resetting its timeout.
3109 This is not a problem when such interrupts are rare. Somebody set up a cron
3110 job to run exiwhat every 2 minutes, which is less than the normal select()
3111 timeout (5 or 10 minutes). This meant that the select() timeout never
3112 kicked in because it was always reset. I have fixed this by comparing the
3113 time when an interrupt arrives with the time at the start of the first call
3114 to select(). If more time than the timeout has elapsed, the interrupt is
3115 treated as a timeout.
3117 5. Some internal re-factoring in preparation for the addition of Sieve
3118 extensions (by MH). In particular, the "personal" test is moved to a
3119 separate function, and given an option for scanning Cc: and Bcc: (which is
3120 not set for Exim filters).
3122 6. When Exim created an email address using the login of the caller as the
3123 local part (e.g. when creating a From: or Sender: header line), it was not
3124 quoting the local part when it contained special characters such as @.
3126 7. Installed new OpenBSD configuration files.
3128 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3129 try to make them clearer.
3131 9. Callout options, other than the timeout value, were being ignored when
3132 verifying sender addresses in header lines. For example, when using
3134 verify = header_sender/callout=no_cache
3136 the cache was (incorrectly) being used.
3138 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3139 only the Cygwin environment.
3141 11. When return_path_on_delivery was set as a log selector, if different remote
3142 addresses in the same message used different return paths and parallel
3143 remote delivery occurred, the wrong values would sometimes be logged.
3144 (Whenever a remote delivery process finished, the return path value from
3145 the most recently started remote delivery process was logged.)
3147 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3148 lines when AUTH and/or TLS are in use. This is the "received protocol"
3149 field. Exim used to use "asmtp" for authenticated SMTP, without any
3150 indication (in the protocol name) for TLS use. Now it follows the RFC and
3151 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3152 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3153 names appear in log lines as well as in Received: header lines.
3155 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3156 extensions, and comparison tests, and to fix some bugs.
3158 14. Changes to the "personal" filter test:
3160 (1) The test was buggy in that it was just doing the equivalent of
3161 "contains" tests on header lines. For example, if a user's address was
3162 anne@some.where, the "personal" test would incorrectly be true for
3164 To: susanne@some.where
3166 This test is now done by extracting each address from the header in turn,
3167 and checking the entire address. Other tests that are part of "personal"
3168 are now done using regular expressions (for example, to check local parts
3169 of addresses in From: header lines).
3171 (2) The list of non-personal local parts in From: addresses has been
3172 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3173 taken from the Sieve specification recommendations.
3175 (3) If the message contains any header line starting with "List-" it is
3176 treated as non-personal.
3178 (4) The test for "circular" in the Subject: header line has been removed
3179 because it now seems ill-conceived.
3181 15. Minor typos in src/EDITME comments corrected.
3183 16. Installed latest exipick from John Jetmore.
3185 17. If headers_add on a router specified a text string that was too long for
3186 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3187 of string_sprintf() is now avoided.
3189 18. $message_body_size was not set (it was always zero) when running the DATA
3190 ACL and the local_scan() function.
3192 19. For the "mail" command in an Exim filter, no default was being set for
3193 the once_repeat time, causing a random time value to be used if "once" was
3194 specified. (If the value happened to be <= 0, no repeat happened.) The
3195 default is now 0s, meaning "never repeat". The "vacation" command was OK
3196 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3197 (I found it when inspecting the code).
3199 20. There is now an overall timeout for performing a callout verification. It
3200 defaults to 4 times the callout timeout, which applies to individual SMTP
3201 commands during the callout. The overall timeout applies when there is more
3202 than one host that can be tried. The timeout is checked before trying the
3203 next host. This prevents very long delays if there are a large number of
3204 hosts and all are timing out (e.g. when the network connections are timing
3205 out). The value of the overall timeout can be changed by specifying an
3206 additional sub-option for "callout", called "maxwait". For example:
3208 verify = sender/callout=5s,maxwait=20s
3210 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3211 to the end before writing, but this should make it even safer).
3213 22. Exim was forgetting that it had advertised PIPELINING for the second and
3214 subsequent messages on an SMTP connection. It was also not resetting its
3215 memory on STARTTLS and an internal HELO.
3217 23. When Exim logs an SMTP synchronization error within a session, it now
3218 records whether PIPELINING has been advertised or not.
3220 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3221 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3222 rather than long int.
3224 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3226 26. Added the never_mail option to autoreply.
3232 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3233 crash if the getsockname() call failed; this can happen if a connection is
3234 closed very soon after it is established. The problem was simply in the
3235 order in which certain operations were done, causing Exim to try to write
3236 to the SMTP stream before it had set up the file descriptor. The bug has
3237 been fixed by making things happen in the correct order.
3243 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3244 before the connection was closed, thus losing the rejection response.
3246 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3247 some early Solaris releases, but causes trouble in current releases where
3248 socklen_t is defined.
3250 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3253 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3256 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3257 that releasing the top of it at the end releases what was used for sub-
3258 expansions (unless the block got too big). However, discard this block if
3259 the first thing is a variable or header, so that we can use its block when
3260 it is dynamic (useful for very large $message_headers, for example).
3262 6. Lookups now cache *every* query, not just the most recent. A new, separate
3263 store pool is used for this. It can be recovered when all lookup caches are
3264 flushed. Lookups now release memory at the end of their result strings.
3265 This has involved some general refactoring of the lookup sources.
3267 7. Some code has been added to the store_xxx() functions to reduce the amount
3268 of flapping under certain conditions.
3270 8. log_incoming_interface used to affect only the <= reception log lines. Now
3271 it causes the local interface and port to be added to several more SMTP log
3272 lines, for example "SMTP connection from", and rejection lines.
3274 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3276 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3278 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3279 could be overwritten at the end of the current message (or the start of a
3280 new message if it was set in a HELO ACL). The value is now preserved for
3281 the duration of the SMTP connection.
3283 12. If a transport had a headers_rewrite setting, and a matching header line
3284 contained an unqualified address, that address was qualified, even if it
3285 did not match any rewriting rules. The underlying bug was that the values
3286 of the flags that permit the existence of unqualified sender and recipient
3287 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3288 non-local messages, and by -bnq for local messages) were not being
3289 preserved with the message after it was received.
3291 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3292 "next input=" as part of the text comprising the host identity instead of
3293 the correct text. The code was using the same buffer for two different
3294 strings. However, depending on which order the printing function evaluated
3295 its arguments, the bug did not always show up. Under Linux, for example, my
3296 test suite worked just fine.
3298 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3299 doesn't work with some older versions of Perl. It has been changed to "my",
3300 which in any case is probably the better facility to use.
3302 15. A really picky compiler found some instances of statements for creating
3303 error messages that either had too many or two few arguments for the format
3306 16. The size of the buffer for calls to the DNS resolver has been increased
3307 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3308 for addresses that have a lot of PTR records. This alleviates a problem; it
3309 does not fully solve it.
3311 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3312 buffer now truncates the list and logs the incident, which is the same
3313 action as happens when Exim is looking up a host name and its aliases.
3314 Previously in this situation something unpredictable would happen;
3315 sometimes it was "internal error: store_reset failed".
3317 18. If a server dropped the connection unexpectedly when an Exim client was
3318 using GnuTLS and trying to read a response, the client delivery process
3319 crashed while trying to generate an error log message.
3321 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3322 string, for example:
3324 warn message = H1: something\nH2: something
3326 the text was added as a single header line from Exim's point of view
3327 though it ended up OK in the delivered message. However, searching for the
3328 second and subsequent header lines using $h_h2: did not work. This has been
3329 fixed. Similarly, if a system filter added multiple headers in this way,
3330 the routers could not see them.
3332 20. Expanded the error message when iplsearch is called with an invalid key to
3333 suggest using net-iplsearch in a host list.
3335 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3336 ACLs are no longer actually imposed (and a message to that effect is
3339 22. If a "gecos" field in a passwd entry contained escaped characters, in
3340 particular, if it contained a \" sequence, Exim got it wrong when building
3341 a From: or a Sender: header from that name. A second bug also caused
3342 incorrect handling when an unquoted " was present following a character
3343 that needed quoting.
3345 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3346 was not being matched caselessly.
3348 24. Arranged for all hyphens in the exim.8 source to be escaped with
3351 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3352 properly. Recipient callout cache records were still being keyed to include
3353 the sender, even when use_sender was set false. This led to far more
3354 callouts that were necessary. The sender is no longer included in the key
3355 when use_sender is false.
3357 26. Added "control = submission" modifier to ACLs.
3359 27. Added the ${base62d: operator to decode base 62 numbers.
3361 28. dnsdb lookups can now access SRV records.
3363 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3364 the configuration file.
3366 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3367 (-v) mode. This makes the output for a verbose queue run more intelligible.
3369 31. Added a use_postmaster feature to recipient callouts.
3371 32. Added the $body_zerocount variable, containing the number of binary zero
3372 bytes in the message body.
3374 33. The time of last modification of the "new" subdirectory is now used as the
3375 "mailbox time last read" when there is a quota error for a maildir
3378 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3380 35. Added +ignore_unknown as a special item in host lists.
3382 36. Code for decoding IPv6 addresses in host lists is now included, even if
3383 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3384 address was recognized as an IP address, but was then not correctly decoded
3385 into binary, causing unexpected and incorrect effects when compared with
3392 1. Very minor rewording of debugging text in manualroute to say "list of
3393 hosts" instead of "hostlist".
3395 2. If verify=header_syntax was set, and a header line with an unqualified
3396 address (no domain) and a large number of spaces between the end of the
3397 name and the colon was received, the reception process suffered a buffer
3398 overflow, and (when I tested it) crashed. This was caused by some obsolete
3399 code that should have been removed. The fix is to remove it!
3401 3. When running in the test harness, delay a bit after writing a bounce
3402 message to get a bit more predictability in the log output.
3404 4. Added a call to search_tidyup() just before forking a reception process. In
3405 theory, someone could use a lookup in the expansion of smtp_accept_max_
3406 per_host which, without the tidyup, could leave open a database connection.
3408 5. Added the variables $recipient_data and $sender_data which get set from a
3409 lookup success in an ACL "recipients" or "senders" condition, or a router
3410 "senders" option, similar to $domain_data and $local_part_data.
3412 6. Moved the writing of debug_print from before to after the "senders" test
3415 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3416 problems for message scanning, either using a data ACL, or using
3417 local_scan() because the Received: header was not generated till after they
3418 were called (in order to set the time as the time of reception completion).
3419 I have revised the way this works. The header is now generated after the
3420 body is received, but before the ACL or local_scan() are called. After they
3421 are run, the timestamp in the header is updated.
3427 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3428 before starting a queue runner without re-exec. This happened only when
3429 deliver_drop_privilege was set or when the Exim user was set to root. The
3430 effect of the bug was that timeouts during subsequent deliveries caused
3431 crashes instead of being properly handled. The handler is now left at its
3432 default (and expected) setting.
3434 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3435 message, again when deliver_drop_privilege is set or Exim is run as root.
3436 The bug described in (1) was not present in this case, but the tidying up
3437 of the other signals was missing. I have made the two cases consistent.
3439 3. The ignore_target_hosts setting on a manualroute router was being ignored
3440 for hosts that were looked up using the /MX notation.
3442 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3445 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3446 operated on the sender address. After changing the $sender_address to <>
3447 for the sender address verify, Exim was re-instated it as the original
3448 (before rewriting) address, but remembering that it had rewritten it, so it
3449 wasn't rewriting it again. This bug also had the effect of breaking the
3450 sender address verification caching when the sender address was rewritten.
3452 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3453 This has been changed so that if the ip literal address matches
3454 ignore_target_hosts, the router declines.
3456 7. Added expansion conditions match_domain, match_address, and match_local_
3457 part (NOT match_host).
3459 8. The placeholder for the Received: header didn't have a length field set.
3461 9. Added code to Exim itself and to exim_lock to test for a specific race
3462 condition that could lead to file corruption when using MBX delivery. The
3463 issue is with the lockfile that is created in /tmp. If this file is removed
3464 after a process has opened it but before that process has acquired a lock,
3465 there is the potential for a second process to recreate the file and also
3466 acquire a lock. This could lead to two Exim processes writing to the file
3467 at the same time. The added code performs the same test as UW imapd; it
3468 checks after acquiring the lock that its file descriptor still refers to
3469 the same named file.
3471 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3472 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3475 11. Added the smtp_active_hostname option. If used, this will typically be made
3476 to depend on the incoming interface address. Because $interface_address is
3477 not set up until the daemon has forked a reception process, error responses
3478 that can happen earlier (such as "too many connections") no longer contain
3481 12. If an expansion in a condition on a "warn" statement fails because a lookup
3482 defers, the "warn" statement is abandoned, and the next ACL statement is
3483 processed. Previously this caused the whole ACL to be aborted.
3485 13. Added the iplsearch lookup type.
3487 14. Added ident_timeout as a log selector.
3489 15. Added tls_certificate_verified as a log selector.
3491 16. Added a global option tls_require_ciphers (compare the smtp transport
3492 option of the same name). This controls incoming TLS connections.
3494 17. I finally figured out how to make tls_require_ciphers do a similar thing
3495 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3496 before starting the TLS session.
3498 18. Tabs are now shown as \t in -bP output.
3500 19. If the log selector return_path_on_delivery was set, Exim crashed when
3501 bouncing a message because it had too many Received: header lines.
3503 20. If two routers both had headers_remove settings, and the first one included
3504 a superfluous trailing colon, the final name in the first list and the
3505 first name in the second list were incorrectly joined into one item (with a
3506 colon in the middle).
3512 1. Added -C and -D options to the exinext utility, mainly to make it easier
3513 to include in the automated testing, but these could be helpful when
3514 multiple configurations are in use.
3516 2. The exinext utility was not formatting the output nicely when there was
3517 an alternate port involved in the retry record key, nor when there was a
3518 message id as well (for retries that were specific to a specific message
3519 and a specific host). It was also confused by IPv6 addresses, because of
3520 the additional colons they contain. I have fixed the IPv4 problem, and
3521 patched it up to do a reasonable job for IPv6.
3523 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3524 delivery, the log line now contains "pipelined" if PIPELINING was used.
3526 4. An SMTP transport process used to panic and die if the bind() call to set
3527 an explicit outgoing interface failed. This has been changed; it is now
3528 treated in the same way as a connect() failure.
3530 5. A reference to $sender_host_name in the part of a conditional expansion
3531 that was being skipped was still causing a DNS lookup. This no longer
3534 6. The def: expansion condition was not recognizing references to header lines
3535 that used bh_ and bheader_.
3537 7. Added the _cache feature to named lists.
3539 8. The code for checking quota_filecount in the appendfile transport was
3540 allowing one more file than it should have been.
3542 9. For compatibility with Sendmail, the command line option
3550 and sets the incoming protocol and host name (for trusted callers). The
3551 host name and its colon can be omitted when only the protocol is to be set.
3552 Note the Exim already has two private options, -pd and -ps, that refer to
3553 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3554 "s", but I don't think that's a major issue.
3556 10. A number of refactoring changes to the code, none of which should affect
3559 (a) The number of logging options was getting close to filling up the
3560 32-bit word that was used as a bit map. I have split them into two classes:
3561 those that are passed in the argument to log_write(), and those that are
3562 only ever tested independently outside of that function. These are now in
3563 separate 32-bit words, so there is plenty of room for expansion again.
3564 There is no change in the user interface or the logging behaviour.
3566 (b) When building, for example, log lines, the code previously used a
3567 macro that called string_cat() twice, in order to add two strings. This is
3568 not really sufficiently general. Furthermore, there was one instance where
3569 it was actually wrong because one of the argument was used twice, and in
3570 one call a function was used. (As it happened, calling the function twice
3571 did not affect the overall behaviour.) The macro has been replaced by a
3572 function that can join an arbitrary number of extra strings onto a growing
3575 (c) The code for expansion conditions now uses a table and a binary chop
3576 instead of a serial search (which was left over from when there were very
3577 few conditions). Also, it now recognizes conditions like "pam" even when
3578 the relevant support is not compiled in: a suitably worded error message is
3579 given if an attempt is made to use such a condition.
3581 11. Added ${time_interval:xxxxx}.
3583 12. A bug was causing one of the ddress fields not to be passed back correctly
3584 from remote delivery subprocesses. The field in question was not being
3585 subsequently used, so this caused to problems in practice.
3587 13. Added new log selectors queue_time and deliver_time.
3589 14. Might have fixed a bug in maildirsizefile handling that threw up
3590 "unexpected character" debug warnings, and recalculated the data
3591 unnecessarily. In any case, I expanded the warning message to give more
3594 15. Added the message "Restricted characters in address" to the statements in
3595 the default ACL that block characters like @ and % in local parts.
3597 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3598 Three changes have been made:
3600 (a) There was a serious bug; a negative response to MAIL caused the whole
3601 recipient domain to be cached as invalid, thereby blocking all messages
3602 to all local parts at the same domain, from all senders. This bug has
3603 been fixed. The domain is no longer cached after a negative response to
3604 MAIL if the sender used is not empty.
3606 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3609 (c) A new callout option, "use_sender" has been added for people who want
3610 the modified behaviour.
3616 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3619 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3620 indeed breaks things for older releases.
3622 3. Added additional logging to the case where there is a problem reading data
3623 from a filter that is running in a subprocess using a pipe, in order to
3624 try to track down a specific problem.
3626 4. Testing facility fudge: when running in the test harness and attempting
3627 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3628 getting "No route to host". Convert this to a timeout.
3630 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3633 6. Some OS don't have socklen_t but use size_t instead. This affects the
3634 fifth argument of getsockopt() amongst other things. This is now
3635 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3636 can be set for individual OS. I have set it for SunOS5, OSF1, and
3637 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3638 some earlier ones do not.
3640 7. Change 4.30/15 was not doing the test caselessly.
3642 8. The standard form for an IPv6 address literal was being rejected by address
3643 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3644 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3645 this, as well as the form without the "IPv6" on the front (but only when
3646 address literals are enabled, of course).
3648 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3650 10. Exim crashed if a message with an empty sender address specified by -f
3651 encountered a router with an errors_to setting. This could be provoked only
3652 by a command such as
3656 where an empty string was supplied; "<>" did not hit this bug.
3658 11. Installed PCRE release 4.5.
3660 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3661 remained set. It is now erased.
3663 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3664 times from message ids (which are base 36 rather than the normal 62).
3666 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3667 were being counted as actual protocol errors, and logged if the log
3668 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3669 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3670 and DATA following a set of rejected RCPTs do not count as protocol errors.
3671 In other words, Exim assumes they were pipelined, though this may not
3672 actually be the case. Of course, in all cases the client gets an
3673 appropriate error code.
3675 15. If a lookup fails in an ACL condition, a message about the failure may
3676 be available; it is used if testing the ACL cannot continue, because most
3677 such messages specify what the cause of the deferral is. However, some
3678 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3679 that caused an old message to be retained and used if a later statement
3680 caused a defer, replacing the real cause of the deferral.
3682 16. If an IP address had so many PTR records that the DNS lookup buffer
3683 was not large enough to hold them, Exim could crash while trying to process
3684 the truncated data. It now detects and logs this case.
3686 17. Further to 4.21/58, another change has been made: if (and only if) the
3687 first line of a message (the first header line) ends with CRLF, a bare LF
3688 in a subsequent header line has a space inserted after it, so as not to
3689 terminate the header.
3691 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3692 unnecessarily, used atoi() instead of strtol(), and didn't check the
3693 termination when getting file sizes from file names by regex.
3695 19. Completely re-implemented the support for maildirsize files, in the light
3696 of a number of problems with the previous contributed implementation
3697 (4.30/29). In particular:
3699 . If the quota is zero, the maildirsize file is maintained, but no quota is
3702 . If the maildir directory does not exist, it is created before any attempt
3703 to write a maildirsize file.
3705 . The quota value in the file is just a cache; if the quota is changed in
3706 the transport, the new value overrides.
3708 . A regular expression is available for excluding directories from the
3711 20. The autoreply transport checks the characters in options that define the
3712 message's headers; it allows continued headers, but it was checking with
3713 isspace() after an embedded newline instead of explicitly looking for a
3716 21. If all the "regular" hosts to which an address was routed had passed their
3717 expiry times, and had not reached their retry times, the address was
3718 bounced, even if fallback hosts were defined. Now Exim should go on to try
3721 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3722 equivalent code in the SMTP transport. Some hosts send humungous responses
3723 to HELO/EHLO, more than 1024 it seems.
3725 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3726 gives compiler warnings in some environments. I've now done it "properly",
3729 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3730 (because of problems with the built-in one) was declared to return uschar *
3731 instead of char *, causing compiler failure.
3733 25. Fixed a file descriptor leak when processing alias/forward files.
3735 26. Fixed a minor format string issue in dbfn.c.
3737 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3739 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3740 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3743 29. When a sender address is verified, it is cached, to save repeating the test
3744 when there is more than one recipient in a message. However, when the
3745 verification involves a callout, it is possible for different callout
3746 options to be set for different recipients. It is too complicated to keep
3747 track of this in the cache, so now Exim always runs a verification when a
3748 callout is required, relying on the callout cache for the optimization.
3749 The overhead is duplication of the address routing, but this should not be
3752 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3753 to be verified with callout=postmaster, and the main callout worked but the
3754 postmaster check failed, the verification correctly failed. However, if a
3755 subsequent RCPT command asked for sender verification *without* the
3756 postmaster check, incorrect caching caused this verification also to fail,
3759 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3760 it was not caching the DNS options (qualify_single, search_parents) that
3761 were used when the lookup failed. A subsequent lookup with different
3762 options therefore always gave the same answer, though there were cases
3763 where it should not have. (Example: a "domains = !$mx_any" option on a
3764 dnslookup router: the "domains" option is always processed without any
3765 widening, but the router might have qualify_single set.) Now Exim uses the
3766 cached value only when the same options are set.
3768 32. Added John Jetmore's "exipick" utility to the distribution.
3770 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3771 than a timeout (e.g. a certificate is required, and is not provided), an
3772 Exim server now closes the connection immediately. Previously it waited for
3773 the client to close - but if the client is SSL, it seems that they each
3774 wait for each other, leading to a delay before one of them times out.
3776 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3777 maintained 0.8.x compatibility because I don't think many are using it, and
3778 it is clearly obsolete.
3780 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3783 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3784 client certificate was expired. A simple patch fixes this, though I don't
3785 understand the full logic of why the verify callback is called multiple
3788 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3789 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3790 which causes problems with some clients (such as the Certicom SSL Plus
3791 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3792 disables the coutermeasure allowing Eudora to connect."
3794 38. Exim was not checking that a write() to a log file succeeded. This could
3795 lead to Bad Things if a log got too big, in particular if it hit a file
3796 size limit. Exim now panics and dies if it cannot write to a log file, just
3797 as it does if it cannot open a log file.
3799 39. Modified OS/Makefile-Linux so that it now contains
3801 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3803 The two -D definitions ensure that Exim is compiled with large file
3804 support, which makes it possible to handle log files that are bigger than
3807 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3808 instance) a domain was checked against a named list that involved a lookup,
3809 causing $domain_data to be set, then another domain was checked against the
3810 same list, then the first domain was re-checked, the value of $domain_data
3811 after the final check could be wrong. In particular, if the second check
3812 failed, it could be set empty. This bug probably also applied to
3815 41. The strip_trailing_dot option was not being applied to the address given
3816 with the -f command-line option.
3818 42. The code for reading a message's header from the spool was incrementing
3819 $received_count, but never initializing it. This meant that the value was
3820 incorrect (doubled) while delivering a message in the same process in which
3821 it was received. In the most common configuration of Exim, this never
3822 happens - a fresh exec is done - but it can happen when
3823 deliver_drop_privilege is set.
3825 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3826 it now includes up to 150 characters of the unexpected data in the log
3829 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3830 and building data strings. The size of both of these buffers was 10 000
3831 bytes - far larger than anybody would *ever* want, thought I. Needless to
3832 say, somebody hit the limit. I have increased the maximum line length to
3833 20 000 and the maximum data length of concatenated lines to 100 000. I have
3834 also fixed two bugs, because there was no checking on these buffers. Tsk,
3835 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3836 buffer is too small.
3838 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3839 lsearch lookups. Now it does.
3841 46. When parsing a route_list item in a manualroute router, a fixed-length
3842 buffer was used for the list of hosts. I made this 1024 bytes long,
3843 thinking that nobody would ever have a list of hosts that long. Wrong.
3844 Somebody had a whole pile of complicated expansion conditions, and the
3845 string was silently truncated, leading to an expansion error. It turns out
3846 that it is easier to change to an unlimited length (owing to other changes
3847 that have happened since this code was originally written) than to build
3848 structure for giving a limitation error. The length of the item that
3849 expands into the list of hosts is now unlimited.
3851 47. The lsearch lookup could not handle data where the length of text line was
3852 more than 4095 characters. Such lines were truncated, leading to shortened
3853 data being returned. It should now handle lines of any length.
3855 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3856 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3859 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3860 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3861 no longer happens in comment lines. A list of replacements is now placed
3862 at the head of all of the source files, except those whose only change is
3863 to replace PERL_COMMAND in the very first #! line.
3865 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3866 messages on the queue, with a bottom-up merge sort, using code contributed
3867 by Michael Haardt. This should make operations like -bp somewhat faster on
3868 large queues. It won't affect queue runners, except when queue_run_in_order
3871 51. Installed eximstats 1.31 in the distribution.
3873 52. Added support for SRV lookups to the dnslookup router.
3875 53. If an ACL referred to $message_body or $message_body_end, the value was not
3876 reset for any messages that followed in the same SMTP session.
3878 54. The store-handling optimization for building very long strings was not
3879 differentiating between the different store pools. I don't think this
3880 actually made any difference in practice, but I've tidied it.
3882 55. While running the routers to verify a sender address, $sender_address
3883 was still set to the sender address. This is wrong, because when routing to
3884 send a bounce to the sender, it would be empty. Therefore, I have changed
3885 it so that, while verifying a sender address, $sender_address is set to <>.
3886 (There is no change to what happens when verifying a recipient address.)
3888 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3889 target A or AAAA records (if not already returned) without resetting the
3890 qualify_single or search_parents options of the DNS resolver. These are
3891 inappropriate in this case because the targets of MX and SRV records must
3892 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3893 target that, when qualified, matched something in the local domain. These
3894 two options are now turned off when doing these lookups.
3896 57. It seems that at least some releases of Reiserfs (which does not have the
3897 concept of a fixed number of inodes) returns zero and not -1 for the
3898 number of available inodes. This interacted badly with check_spool_inodes,
3899 which assumed that -1 was the "no such thing" setting. What I have done is
3900 to check that the total number of inodes is greater than zero before doing
3901 the test of how many are available.
3903 58. When a "warn" ACL statement has a log_message modifier, the message is
3904 remembered, and not repeated. This is to avoid a lot of repetition when a
3905 message has many recipients that cause the same warning to be written.
3906 Howewer, Exim was preserving the list of already written lines for an
3907 entire SMTP session, which doesn't seem right. The memory is now reset if a
3908 new message is started.
3910 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3911 debugging output unless log_rewrite was also set.
3913 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3914 of (int)(handle) when we know that handle contains (void *)(-1).
3916 61. The Exim daemon panic-logs an error return when it closes the incoming
3917 connection. However "connection reset by peer" seems to be common, and
3918 isn't really an error worthy of noting specially, so that particular error
3921 62. When Exim is trying to find all the local interfaces, it used to panic and
3922 die if the ioctl to get the interface flags failed. However, it seems that
3923 on at least one OS (Solaris 9) it is possible to have an interface that is
3924 included in the list of interfaces, but for which you get a failure error
3925 for this call. This happens when the interface is not "plumbed" into a
3926 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3927 failure of the "get flags" call assumes that the interface is down.
3929 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3930 makes life easier for people who are doing arithmetic on fields extracted
3931 from dates, where you often get leading zeros that should not be
3932 interpreted as octal.
3934 64. Added qualify_domain to the redirect router, to override the global
3937 65. If a pathologically long header line contained very many addresses (the
3938 report of this problem mentioned 10 000) and each of them was rewritten,
3939 Exim could use up a very large amount of memory. (It kept on making new
3940 copies of the header line as it rewrote, and never released the old ones.)
3941 At the expense of a bit more processing, the header rewriting function has
3942 been changed so that it no longer eats memory in this way.
3944 66. The generation of the Received: header has been moved from the time that a
3945 message starts to be received, to the time that it finishes. The timestamp
3946 in the Received: header should now be very close to that of the <= log
3947 line. There are two side-effects of this change:
3949 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3950 logged header lines no longer include the local Received: line, because
3951 it has not yet been created. The same applies to a copy of the message
3952 that is returned to a non-SMTP sender when a message is rejected.
3954 (b) When a filter file is tested using -bf, no additional Received: header
3955 is added to the test message. After some thought, I decided that this
3958 This change does not affect the value of $received_for. It is still set
3959 after address rewriting, but before local_scan() is called.
3961 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3963 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3964 gave an unhelpful panic error message, and a defer error. I have managed to
3965 change this behaviour so that it now rejects any supplied certificate,
3966 which seems right, as the list of acceptable certificates is empty.
3968 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3969 gave an unhelpful defer error. I have not managed to make this reject any
3970 supplied certificates, but the error message it gives is "no certificate
3971 supplied", which is not helpful.
3973 70. exigrep's output now also includes lines that are not associated with any
3974 message, but which match the given pattern. Implemented by a patch from
3975 Martin Sluka, which also tidied up the Perl a bit.
3977 71. Recipient callout verification, like sender verification, was using <> in
3978 the MAIL FROM command. This isn't really the right thing, since the actual
3979 sender may affect whether the remote host accepts the recipient or not. I
3980 have changed it to use the actual sender in the callout; this means that
3981 the cache record is now keyed on a recipient/sender pair, not just the
3982 recipient address. There doesn't seem to be a real danger of callout loops,
3983 since a callout by the remote host to check the sender would use <>.
3984 [SEE ABOVE: changed after hitting problems.]
3986 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3987 temporary errors. However, in the case of such a code being given after
3988 the end of a data transmission (i.e. after ".") Exim was failing to write
3989 a retry record for the message. (Yes, there was some broken host that was
3990 actually sending 8xx at this point.)
3992 73. An unknown lookup type in a host list could cause Exim to panic-die when
3993 the list was checked. (An example that provoked this was putting <; in the
3994 middle of a list instead of at the start.) If this happened during a DATA
3995 ACL check, a -D file could be left lying around. This kind of configuration
3996 error no longer causes Exim to die; instead it causes a defer errror. The
3997 incident is still logged to the main and panic logs.
3999 74. Buglet left over from Exim 3 conversion. The message "too many messages
4000 in one connection" was written to the rejectlog but not the mainlog, except
4001 when address rewriting (yes!) was being logged.
4003 75. Added write_rejectlog option.
4005 76. When a system filter was run not as root (that is, when system_filter_user
4006 was set), the values of the $n variables were not being returned to the
4007 main process; thus, they were not subsequently available in the $sn
4010 77. Added +return_path_on_delivery log selector.
4012 78. A connection timeout was being treated differently from recipients deferred
4013 when testing hosts_max_try with a message that was older than the host's
4014 retry timeout. (The host should not be counted, thus allowing all hosts to
4015 be tried at least once before bouncing.) This may have been the cause of an
4016 occasionally reported bug whereby a message would remain on the queue
4017 longer than the retry timeout, but would be bounced if a delivery was
4018 forced. I say "may" because I never totally pinned down the problem;
4019 setting up timeout/retry tests is difficult. See also the next item.
4021 79. The ultimate address timeout was not being applied to errors that involved
4022 a combination of host plus message (for example, a timeout on a MAIL
4023 command). When an address resolved to a number of possible hosts, and they
4024 were not all tried for each delivery (e.g. because of hosts_max_try), a
4025 message could remain on the queue longer than the retry timeout.
4027 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4030 81. Fixed an obscure SMTP outgoing bug which required at least the following
4031 conditions: (a) there was another message waiting for the same server;
4032 (b) the server returned 5xx to all RCPT commands in the first message so
4033 that the message was not completed; (c) the server dropped the connection
4034 or gave a negative response to the RSET that Exim sends to abort the
4035 transaction. The observed case was a dropped connection after DATA that had
4036 been sent in pipelining mode. That is, the server had advertised PIPELINING
4037 but was not implementing it correctly. The effect of the bug was incorrect
4038 behaviour, such as trying another host, and this could lead to a crash.
4044 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4045 and daemon.c were passed as pointers to ints; they should have been
4046 pointers to socklen_t variables (which are typically unsigned ints).
4048 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4051 3. Fixed a really odd bug that affected only the testing scheme; patching a
4052 certain fixed string in the binary changed the value of another string that
4053 happened to be identical to the end of the original first string.
4055 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4056 name", it returns that address as the IP address. On some operating
4057 systems (e.g. Solaris), it also passes back the IP address string as the
4058 "host name". However, on others (e.g. Linux), it passes back an empty
4059 string. Exim wasn't checking for this, and was changing the host name to an
4060 empty string, assuming it had been canonicized.
4062 5. Although rare, it is permitted to have more than one PTR record for a given
4063 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4064 all the names associated with an address, because they do in Solaris.
4065 However, it seems that they do not in Linux for data that comes from the
4066 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4067 I found this out when I moved to a new Linux workstation and tried to run
4068 the Exim test suite.
4070 To get round this problem I have changed the code so that it now does its
4071 own call to the DNS to look up PTR records when searching for a host name.
4072 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4073 addresses that are only in /etc/hosts are still found.
4075 This behaviour is, however, controlled by an option called host_lookup_
4076 order, which defaults to "bydns:byaddr". If people want to use the other
4077 order, or indeed, just use one or the other means of lookup, they can
4078 specify it in this variable.
4080 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4081 some operating systems, this comes back from gethostbyaddr() as an empty
4082 string, and this is what Exim used to test for. However, it seems that in
4083 other systems, "." is yielded. Exim now tests for this case too.
4085 7. The values of check_spool_space and check_log_space are now held internally
4086 as a number of kilobytes instead of an absolute number of bytes. If a
4087 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4088 kilobyte. This means that much larger values can be stored.
4090 8. Exim monitor: an attempt to get the action menu when not actually pointing
4091 at a message produces an empty menu entitled "No message selected". This
4092 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4093 no entries in it ("Shell widget menu has zero width and/or height"). So I
4094 have added a single, blank menu entry in this case.
4096 9. Added ${quote_local_part.
4098 10. MIME decoding is now applied to the contents of Subject: header lines when
4101 11. Now that a reference to $sender_host_address automatically causes a reverse
4102 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4103 host lookup before query-style lookups in lists that might use this
4104 variable. This has therefore been abolished, and the "net-" prefix is no
4105 longer necessary for query-style lookups.
4107 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4108 have been a typo for LFLAGS, so it has been changed.
4110 13. The install script calls Exim with "-C /dev/null" in order to find the
4111 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4112 to be output. Howeve, since Exim outputs its version number before the
4113 error, it didn't break the script. It just looked ugly. I fixed this by
4114 always allowing "-C /dev/null" if the caller is root.
4116 14. Ignore overlarge ACL variable number when reading spool file - insurance
4117 against a later release with more variables having written the file.
4119 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4120 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4121 this, as well as the form without the "IPv6" on the front.
4123 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4124 OS/Makefile-Darwin file.
4126 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4127 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4129 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4130 variable. However, some OS use ints for inodes. I've added cast to long int
4131 to get rid of the compiler warning.
4133 19. I had forgotten to lock out "/../" in configuration file names when
4134 ALT_CONFIG_PREFIX was set.
4136 20. Routers used for verification do not need to specify transports. However,
4137 if such a router generated a host list, and callout was configured, Exim
4138 crashed, because it could not find a port number from the (non-existent)
4139 transport. It now assumes port 25 in this circumstance.
4141 21. Added the -t option to exigrep.
4143 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4144 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4145 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4146 them is set, LOOKUP_LSEARCH is forced.
4148 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4150 24. Added sender and host information to the "rejected by local_scan()" log
4151 line; previously there was no indication of these.
4153 25. Added .include_if_exists.
4155 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4156 Linux. It turns out that not all file systems support this. Apparently some
4157 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4158 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4159 not supported on the file descriptor, is now ignored when Exim is trying to
4160 sync a directory. This applies only to Linux.
4162 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4164 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4165 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4168 29. Added support for maildirsize files from supplied patch (modified a bit).
4170 30. The use of :fail: followed by an empty string could lead Exim to respond to
4171 sender verification failures with (e.g.):
4173 550 Verification failed for <xxx>
4174 550 Sender verify failed
4176 where the first response line was missing the '-' that indicates it is not
4177 the final line of the response.
4179 31. The loop for finding the name of the user that called Exim had a hardwired
4180 limit of 10; it now uses the value of finduser_retries, which is used for
4181 all other user lookups.
4183 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4186 33. Exim was neglecting to zero errno before one call of strtol() when
4187 expanding a string and expecting an integer value. On some systems this
4188 resulted in spurious "integer overflow" errors. Also, it was casting the
4189 result into an int without checking.
4191 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4192 did not work. The code looks as if it has *never* worked, though it appears
4193 to have been documented since at least releast 1.62. I have made it work.
4195 35. The "timeout_DNS" error in retry rules, also documented since at least
4196 1.62, also never worked. As it isn't clear exactly what this means, and
4197 clearly it isn't a major issue, I have abolished the feature by treating it
4198 as "timeout", and writing a warning to the main and panic logs.
4200 36. The display of retry rules for -brt wasn't always showing the error code
4203 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4204 timeout_connect_A, timeout_connect_MX.
4206 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4207 to the empty sender.
4209 39. The daemon was not analysing the content of -oX till after it had closed
4210 stderr and disconnected from the controlling terminal. This meant that any
4211 syntax errors were only noted on the panic log, and the return code from
4212 the command was 0. By re-arranging the code a little, I've made the
4213 decoding happen first, so such errors now appear on stderr, and the return
4214 code is 1. However, the actual setting up of the sockets still happens in
4215 the disconnected process, so errors there are still only recorded on the
4218 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4219 connections (as happens on some IP stacks) was logged at start up time as
4220 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4221 it from "IPv6 and IPv4", which means that two separate sockets are being
4224 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4225 says whether AF_INET or AF_INET6 was passed as an argument.
4227 42. Exiwhat output was messed up when time zones were included in log
4230 43. Exiwhat now gives more information about the daemon's listening ports,
4231 and whether -tls-on-connect was used.
4233 44. The "port" option of the smtp transport is now expanded.
4235 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4236 silently ignored. Now an error message is written to the main and panic
4239 46. There's a new ACL modifier called "logwrite" which writes to a log file
4240 as soon as it is encountered.
4242 47. Added $local_user_uid and $local_user_gid at routing time.
4244 48. Exim crashed when trying to verify a sender address that was being
4247 49. Exim was recognizing only a space character after ".include". It now also
4248 recognizes a tab character.
4250 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4251 extracting the relevant information from the specification. The man page no
4252 longer contains scrambled data for the -d option, and I've added a section
4253 at the front about calling Exim under different names.
4255 51. Added "extra_headers" argument to the "mail" command in filter files.
4257 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4260 53. Installed eximstats 1.29.
4262 54. Added transport_filter_timeout as a generic transport option.
4264 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4265 Cc: header lines. This was required by RFC 822, but it not required by RFC
4268 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4269 incoming messages that did not have them. Now it does so only if the
4270 message originates locally, that is, if there is no associated remote host
4271 address. When Resent- header lines are present, this applies to the Resent-
4272 lines rather than the non-Resent- lines.
4274 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4275 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4277 58. Messages for configuration errors now include the name of the main
4278 configuration file - useful now that there may be more than one file in a
4279 list (.included file names were always shown).
4281 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4282 for those rare installations that do not start the daemon as root or run it
4283 setuid root. I've cut out the call to initgroups() if the daemon is not
4286 60. The Exim user and group can now be bound into the binary as text strings
4287 that are looked up at the start of Exim's processing.
4289 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4291 62. Added $mailstore_basename variable.
4293 63. Installed patch to sieve.c from Michael Haardt.
4295 64. When Exim failed to open the panic log after failing to open the main log,
4296 the original message it was trying to log was written to stderr and debug
4297 output, but if they were not available (the usual case in production), it
4298 was lost. Now it is written to syslog before the two lines that record the
4299 failures to open the logs.
4301 65. Users' Exim filters run in subprocesses under the user's uid. It is
4302 possible for a "deliver" command or an alias in a "personal" command to
4303 provoke an address rewrite. If logging of address rewriting is configured,
4304 this fails because the process is not running as root or exim. There may be
4305 a better way of dealing with this, but for the moment (because 4.30 needs
4306 to be released), I have disabled address rewrite logging when running a
4307 filter in a non-root, non-exim process.
4313 1. The buildconfig auxiliary program wasn't quoting the value set for
4314 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4315 not defined. This bug was masked in 4.22 by the effect that was fixed in
4318 2. Some messages that were rejected after a message id was allocated were
4319 shown as "incomplete" by exigrep. It no longer does this for messages that
4320 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4322 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4323 have allow_domain_literals set, the ID did not get logged in the <= line.
4324 Domain literals are now always recognized in Message-ID: header lines.
4326 4. The first argument for a ${extract expansion item is the key name or field
4327 number. Leading and trailing spaces in this item were not being ignored,
4328 causing some misleading effects.
4330 5. When deliver_drop_privilege was set, single queue runner processes started
4331 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4332 same command in the process it spins off) were not dropping privilege.
4334 6. When the daemon running as "exim" started a queue runner, it always
4335 re-executed Exim in the spun-off process. This is a waste of effort when
4336 deliver_drop_privilege is set. The new process now just calls the
4337 queue-runner function directly.
4343 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4346 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4347 ignored. Though the use of -oP was forcing the writing of a pid file, it
4348 was always written to the default place.
4350 3. If the message "no IP address found for host xxxx" is generated during
4351 incoming verification, it is now followed by identification of the incoming
4352 connection (so you can more easily find what provoked it).
4354 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4356 5. Added some features to "harden" Exim a bit more against certain attacks:
4358 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4359 be put in Local/Makefile. This is like the never_users runtime option,
4360 but it cannot be overridden. The default setting is "root".
4362 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4363 prefix string with which any file named in a -C command line option
4366 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4367 is retained for -C and -D only if the caller of Exim is root. Without
4368 it, the exim user may also use -C and -D and retain privilege.
4370 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4371 command line option is disabled.
4373 6. Macro names set by the -D option must start with an upper case letter, just
4374 like macro names defined in the configuration file.
4376 7. Added "dereference=" facility to LDAP.
4378 8. Two instances of the typo "uknown" in the source files are fixed.
4380 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4381 the Configure-Makefile script screwed up while processing it.
4383 10. Incorporated PCRE 4.4.
4385 11. The SMTP synchronization check was not operating right at the start of an
4386 SMTP session. For example, it could not catch a HELO sent before the client
4387 waited for the greeting. There is now a check for outstanding input at the
4388 point when the greeting is written. Because of the duplex, asynchronous
4389 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4390 way, but not yet received, when the check is performed.
4392 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4393 on TCP/IP sockets, because this apparently causes some broken clients to
4396 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4397 unchanged) from the Cygwin maintainer.
4399 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4400 was supported instead of testing for mbx. Effectively a typo.
4402 15. The spa authenticator server code was not checking that the input it
4403 received was valid base64.
4405 16. The debug output line for the "set" modifier in ACLs was not showing the
4406 name of the variable that was being set.
4408 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4410 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4411 lookup on its own. Something else was needed to trigger the lookup. For
4412 example, a match in host_lookup or the need for a host name in a host list.
4413 Now, if $sender_host_name is referenced and the host name has not yet been
4414 looked up, a lookup is performed. If the lookup fails, the variable remains
4415 empty, and $host_lookup_failed is set to "1".
4417 19. Added "eqi" as a case-independent comparison operator.
4419 20. The saslauthd authentication condition could segfault if neither service
4420 nor realm was specified.
4422 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4423 error message that was logged was misleading, and incoming SMTP
4424 connections were dropped. The message is now more accurate, and temporary
4425 errors are given to SMTP connections.
4427 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4428 (except RSET) with a 421 error, until QUIT is received. However, it was
4429 failing to send a response to QUIT.
4431 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4432 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4433 ACL is now run after the helo_try_verify_hosts code.
4435 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4436 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4437 case-independent, so other case variants are also recognized). Apparently
4438 some systems use these upper case variants.
4440 25. If more than two messages were waiting for the same host, and a transport
4441 filter was specified for the transport, Exim sent two messages over the
4442 same TCP/IP connection, and then failed with "socket operation on non-
4443 socket" when it tried to send the third.
4445 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4447 27. The extern definition of crypt16() in expand.c was not being excluded when
4448 the OS had its own crypt16() function.
4450 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4451 as a preferred synonym for return_size_limit, both as an option and as an
4454 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4456 30. Changed the default configuration ACL to relax the local part checking rule
4457 for addresses that are not in any local domains. For these addresses,
4458 slashes and pipe symbols are allowed within local parts, but the sequence
4459 /../ is explicitly forbidden.
4461 31. SPA server authentication was not clearing the challenge buffer before
4464 32. log_message in a "warn" ACL statement was writing to the reject log as
4465 well as to the main log, which contradicts the documentation and doesn't
4466 seem right (because no rejection is happening). So I have stopped it.
4468 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4469 However, I am unable to do any testing of this.
4471 34. Fixed an infelicity in the appendfile transport. When checking directories
4472 for a mailbox, to see if any needed to be created, it was accidentally
4473 using path names with one or more superfluous leading slashes; tracing
4474 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4476 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4477 contents are added to the log line that is written for every discarded
4478 recipient. (Previously a log_message setting was ignored.)
4480 36. The ${quote: operator now quotes the string if it is empty.
4482 37. The install script runs exim in order to find its version number. If for
4483 some reason other than non-existence or emptiness, which it checks, it
4484 could not run './exim', it was installing it with an empty version number,
4485 i.e. as "exim-". This error state is now caught, and the installation is
4488 38. An argument was missing from the function that creates an error message
4489 when Exim fails to connect to the socket for saslauthd authentication.
4490 This could cause Exim to crash, or give a corrupted message.
4492 39. Added isip, isip4, and isip6 to ${if conditions.
4494 40. The ACL variables $acl_xx are now saved with the message, and can be
4495 accessed later in routers, transports, and filters.
4497 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4498 strings in the file are not string-expanded.
4500 42. If a MAIL command specified a SIZE value that was too large to fit into an
4501 int variable, the check against message_size_limit failed. Such values are
4502 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4503 day this will have to be increased, but I don't think I want to be around
4504 when emails are that large.
4511 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4512 iconv() is not standard in FreeBSD.
4514 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4515 IPv6 enabled. The observed symptom was a segmentation fault on return from
4516 the function os_common_find_running_interfaces() in src/os.c.
4518 3. In the check_special_case() function in daemon.c I had used "errno" as an
4519 argument name, which causes warnings on some systems. This was basically a
4520 typo, since it was named "eno" in the comments!
4522 4. The code that waits for the clock to tick (at a resolution of some fraction
4523 of a second) so as to ensure message-id uniqueness was always waiting for
4524 at least one whole tick, when it could have waited for less. [This is
4525 almost certainly not relevant at current processor speeds, where it is
4526 unlikely to ever wait at all. But we try to future-proof.]
4528 5. The function that sleeps for a time interval that includes fractions of a
4529 second contained a race. It did not block SIGALRM between setting the
4530 timer, and suspending (a couple of lines later). If the interval was short
4531 and the sigsuspend() was delayed until after it had expired, the suspension
4532 never ended. On busy systems this could lead to processes getting stuck for
4535 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4536 process, before it forks any delivery processes. The open lookup caching
4537 mechanism meant that the open file or database connection was passed into
4538 the delivery process. The problem was that delivery processes always tidy
4539 up cached lookup data. This could cause a problem for the next delivery
4540 process started by the queue runner, because the external queue runner
4541 process does not know about the closure. So the next delivery process
4542 still has data in the lookup cache. In the case of a file lookup, there was
4543 no problem because closing a file descriptor in a subprocess doesn't affect
4544 the parent. However, if the lookup was caching a connection to a database,
4545 the connection was closed, and the second delivery process was likely to
4546 see errors such as "PGSQL: query failed: server closed the connection
4547 unexpectedly". The problem has been fixed by closing all cached lookups
4548 in a queue runner before running a delivery process.
4550 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4551 seem to have the "const" qualifier which it has on other OS. I've
4554 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4555 *of the same type* (client or server) with the same public name that an
4556 error should be diagnosed.
4558 9. When Exim looked up a host name for an IP address, but failed to find the
4559 original IP address when looking up the host name (a safety check), it
4560 output the message "<ip address> does not match any IP for NULL", which was
4561 confusing, to say the least. The bug was that the host name should have
4562 appeared instead of "NULL".
4564 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4565 user that is built into the binary, and the -C or -D options is used, root
4566 privilege is dropped before the configuration file is read. In addition,
4567 logging is switched to stderr instead of the normal log files. If the
4568 configuration then re-defines the Exim user, the unprivileged environment
4569 is probably not what is expected, so Exim logs a panic warning message (but
4572 However, if deliver_drop_privilege is set, the unprivileged state may well
4573 be exactly what is intended, so the warning has been cut out in that case,
4574 and Exim is allowed to try to write to its normal log files.
4580 1. smtp_return_error_details was not giving details for temporary sender
4581 or receiver verification errors.
4583 2. Diagnose a configuration error if two authenticators have the same public
4586 3. Exim used not to create the message log file for a message until the first
4587 delivery attempt. This could be confusing when incoming messages were held
4588 for policy or load reasons. The message log file is now created at the time
4589 the message is received, and an initial "Received" line is written to it.
4591 4. The automatically generated man page for command line options had a minor
4592 bug that caused no ill effects; however, a more serious problem was that
4593 the procedure for building the man page automatically didn't always
4594 operate. Consequently, release 4.20 contains an out-of-date version. This
4595 shouldn't happen again.
4597 5. When building Exim with embedded Perl support, the script that builds the
4598 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4599 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4601 6. The freeze_tell option was not being used for messages that were frozen on
4602 arrival, either by an ACL or by local_scan().
4604 7. Added the smtp_incomplete_transaction log selector.
4606 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4607 was accepting AUTH without a new EHLO.
4609 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4610 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4613 10. Logging of TCP/IP connections (when configured) now happens in the main
4614 daemon process instead of the child process, so that the TCP/IP connection
4615 count is more accurate (but it can never be perfect).
4617 11. The use of "drop" in a nested ACL was not being handled correctly in the
4618 outer ACL. Now, if condition failure induced by the nested "drop" causes
4619 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4620 or "require"), the connection is dropped.
4622 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4623 that yield "discard" can now be used with an "accept" or a "discard" verb,
4624 but an error is generated for any others (because I can't see a useful way
4625 to define what should happen).
4627 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4628 are now processed in the same way as lines in the Exim configuration file.
4629 In particular, continuation lines are supported.
4631 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4633 15. Added -ti meaning -t -i.
4635 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4636 domains, and warn by logging if others are found.
4638 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4639 returned from GIFCONF when Exim is trying to find the list of interfaces on
4640 a host. The code in os.c has been modified to copy each ifreq to an aligned
4641 structure in all cases.
4643 Also, in some cases, the returned ifreq's were being copied to a 'struct
4644 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4645 means the last couple of bytes of an IPv6 address could be chopped if the
4646 ifreq contained only a normal sockaddr (14 bytes storage).
4648 18. Named domain lists were not supported in the hosts_treat_as_local option.
4649 An entry such as +xxxx was not recognized, and was treated as a literal
4652 19. Ensure that header lines added by a DATA ACL are included in the reject log
4653 if the ACL subsequently rejects the message.
4655 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4656 MD5 (which is deprecated).
4658 21. When testing a filter file using -bf, Exim was writing a message when it
4659 took the sender from a "From " line in the message, but it was not doing so
4660 when it took $return_path from a Return-Path: header line. It now does.
4662 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4663 with a valid header line field name (a series of printing characters
4664 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4666 23. Changed "disc" in the source to "disk" to conform to the documentation and
4667 the book and for uniformity.
4669 24. Ignore Sendmail's -Ooption=value command line item.
4671 25. When execve() failed while trying to run a command in a pipe transport,
4672 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4673 could be confused with a return value of 69 from the command itself. This
4674 has been changed to 127, the value the shell returns if it is asked to run
4675 a non-existent command. The wording for the related log line suggests a
4676 non-existent command as the problem.
4678 26. If received_header_text expands to an empty string, do not add a Received:
4679 header line to the message. (Well, it adds a token one on the spool, but
4680 marks it "old" so that it doesn't get used or transmitted.)
4682 27. Installed eximstats 1.28 (addition of -nt option).
4684 28. There was no check for failure on the call to getsockname() in the daemon
4685 code. This can fail if there is a shortage of resources on the system, with
4686 ENOMEM, for example. A temporary error is now given on failure.
4688 29. Contrary to the C standard, it seems that in some environments, the
4689 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4690 program. Exim now does this explicitly; it affects the formatting of
4691 timestamps using strftime().
4693 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4694 complaints. I've now initialized all the variables, to avoid this.
4696 32. Header lines added by a system filter were not being "seen" during
4697 transport-time rewrites.
4699 33. The info_callback() function passed to OpenSSL is set up with type void
4700 (*)(SSL *, int, int), as described somewhere. However, when calling the
4701 function (actually a macro) that sets it up, the type void(*)() is
4702 expected. I've put in a cast to prevent warnings from picky compilers.
4704 34. If a DNS black list lookup found a CNAME record, but there were no A
4705 records associated with the domain it pointed at, Exim crashed.
4707 35. If a DNS black list lookup returned more than one A record, Exim ignored
4708 all but the first. It now scans all returned addresses if a particular IP
4709 value is being sought. In this situation, the contents of the
4710 $dnslist_value variable are a list of all the addresses, separated by a
4713 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4714 to accept a host name and all its aliases if the forward lookup for any of
4715 them yielded the IP address of the incoming connection. Now it accepts only
4716 those names whose forward lookup yields the correct IP address. Any other
4717 names are discarded. This closes a loophole whereby a rogue DNS
4718 administrator could create reverse DNS records to break through a
4719 wildcarded host restriction in an ACL.
4721 37. If a user filter or a system filter that ran in a subprocess used any of
4722 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4723 the wrong values were passed to the pipe command ($thisaddress had the
4724 value of $0, $0 had the value of $1, etc). This bug was introduced by
4725 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4727 38. Improved the line breaking for long SMTP error messages from ACLs.
4728 Previously, if there was no break point between 40 and 75 characters, Exim
4729 left the rest of the message alone. Two changes have been made: (a) I've
4730 reduced the minimum length to 35 characters; (b) if it can't find a break
4731 point between 35 and 75 characters, it looks ahead and uses the first one
4732 that it finds. This may give the occasional overlong line, but at least the
4733 remaining text gets split now.
4735 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4736 file descriptors might be low, and that setting 1000 would always raise it.
4737 It turns out that in some environments, the limit is already over 1000 and
4738 that lowering it causes trouble. So now Exim takes care not to decrease it.
4740 40. When delivering a message, the value of $return_path is set to $sender_
4741 address at the start of routing (routers may change the value). By an
4742 oversight, this default was not being set up when an address was tested by
4743 -bt or -bv, which affected the outcome if any router or filter referred to
4746 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4747 log only when "message" or "log_message" are set. However, if one of the
4748 conditions was an address verification, or a call to a nested ACL, the
4749 messages generated by the underlying test were being passed through. This
4750 no longer happens. The underlying message is available in $acl_verify_
4751 message for both "message" and "log_message" expansions, so it can be
4752 passed through if needed.
4754 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4755 new expansion $bh_ to give the encoded byte string without charset
4756 translation. Translation happens only if iconv() is available; HAVE_ICONV
4757 indicates this at build time. HEADERS_CHARSET gives the charset to
4758 translate to; headers_charset can change it in the configuration, and
4759 "headers charset" can change it in an individual filter file.
4761 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4762 that creates RFC 2047 encoded "words" labels them as that charset instead
4763 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4764 expansion operator; (ii) when Exim creates a From: line for a local
4765 message; (iii) when a header line is rewritten to include a "phrase" part.
4767 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4768 buggy, causing it to skip the first lines of messages whose message ID
4769 ended in 'D'. This would not have bitten before Exim release 4.14, because
4770 message IDs were unlikely to end in 'D' before then. The effect was to have
4771 incorrect size information for certain domains.
4773 45. #include "config.h" was missing at the start of the crypt16.c module. This
4774 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4777 46. If there was a timeout during a "random" callout check, Exim treated it as
4778 a failure of the random address, and carried on sending RSET and the real
4779 address. If the delay was just some slowness somewhere, the response to the
4780 original RCPT would be taken as a response to RSET and so on, causing
4781 mayhem of various kinds.
4783 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4784 when I implemented it. It didn't allow for the fact that some option values
4785 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4786 the right test for positive values.
4788 48. Domain names in DNS records are case-independent. Exim always looks them up
4789 in lower case. Some resolvers return domain names in exactly the case they
4790 appear in the zone file, that is, they may contain uppercase letters. Not
4791 all resolvers do this - some return always lower case. Exim was treating a
4792 change of case by a resolver as a change of domain, similar to a widening
4793 of a domain abbreviation. This triggered its re-routing code and so it was
4794 trying to route what was effectively the same domain again. This normally
4795 caused routing to fail (because the router wouldn't handle the domain
4796 twice). Now Exim checks for this case specially, and just changes the
4797 casing of the domain that it ultimately uses when it transmits the message
4800 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4803 50. If a filter generated a file delivery with a non-absolute name (possible if
4804 no home directory exists for the router), the forbid_file option was not
4807 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4808 the existing equality matching.
4810 52. Exim was using ints instead of ino_t variables in some places where it was
4811 dealing with inode numbers.
4813 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4814 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4815 called TMPDIR, and if it finds it is different, it changes its value.
4817 54. The smtp_printf() function is now made available to local_scan() so
4818 additional output lines can be written before returning. There is also an
4819 smtp_fflush() function to enable the detection of a dropped connection.
4820 The variables smtp_input and smtp_batched_input are exported to
4823 55. Changed the default runtime configuration: the message "Unknown user"
4824 has been removed from the ACL, and instead placed on the localuser router,
4825 using the cannot_route_message feature. This means that any verification
4826 failures that generate their own messages won't get overridden. Similarly,
4827 the "Unrouteable address" message that was in the ACL for unverifiable
4828 relay addresses has also been removed.
4830 56. Added hosts_avoid_esmtp to the smtp transport.
4832 57. The exicyclog script was not checking for the esoteric option
4833 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4834 will work only if exicyclog is run under the appropriate euid.
4836 58. Following a discussion on the list, the rules by which Exim recognises line
4837 endings on incoming messages have been changed. The -dropcr and drop_cr
4838 options are now no-ops, retained only for backwards compatibility. The
4839 following line terminators are recognized: LF CRLF CR. However, special
4840 processing applies to CR:
4842 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4843 nor a local message in the state where . is a terminator.
4845 (ii) If a bare CR is encountered in a header line, an extra space is added
4846 after the line terminator so as not to end the header. The reasoning
4847 behind this is that bare CRs in header lines are most likely either
4848 to be mistakes, or people trying to play silly games.
4850 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4851 was being incorrectly given as 18 bytes larger than it should have been.
4852 This is a VOB (very old bug).
4854 60. This may never have affected anything current, but just in case it has:
4855 When the local host is found other than at the start of a list of hosts,
4856 the local host, those with the same MX, and any that follow, are discarded.
4857 When the list in question was part of a longer list of hosts, the following
4858 hosts (not currently being processed) were also being discarded. This no
4859 longer happens. I'm not sure if this situation could ever has previously
4862 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4865 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4866 header. This is something that is recommended in a new Internet Draft, and
4867 is something that is documented as being done by Sendmail. There are two
4868 possible values. For messages generated by the autoreply transport, Exim
4871 Auto-Submitted: auto-replied
4873 whereas for all other generated messages (e.g. bounces) it adds
4875 Auto-Submitted: auto-generated
4877 63. The "personal" condition in filters now includes a test for the
4878 Auto-Submitted: header. If it contains the string "auto-" the message it
4879 not considered personal.
4881 64. Added rcpt_include_affixes as a generic transport option.
4883 65. Added queue_only_override (default true).
4885 66. Added the syslog_duplication option.
4887 67. If what should have been the first header line of a message consisted of
4888 a space followed by a colon, Exim was mis-interpreting it as a header line.
4889 It isn't of course - it is syntactically invalid and should therefore be
4890 treated as the start of the message body. The misbehaviour could have
4891 caused a number of strange effects, including loss of data in subsequent
4892 header lines, and spool format errors.
4894 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4895 client host had authenticated. This control can now be exercised by an ACL
4896 for more flexibility.
4898 69. By default, callouts do not happen when testing with -bh. There is now a
4899 variant, -bhc, which does actually run the callout code, including
4900 consulting and updating the callout cache.
4902 70. Added support for saslauthd authentication, courtesy of Alexander
4905 71. If statvfs() failed on the spool or log directories while checking their
4906 size for availability, Exim confusingly gave the error "space shortage".
4907 Furthermore, in debugging mode it crashed with a floating point exception.
4908 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4909 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4910 serious problem, Exim now writes to the main and panic logs when this
4911 happens, with details of the failure. It then refuses to accept the
4912 incoming message, giving the message "spool directory problem" or "log
4913 directory problem" with a 421 code for SMTP messages.
4915 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4916 0, 1, and 2 exist, because some OS complain for execs without them (see
4917 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4918 descriptors. However, the code omitted to check that the open succeeded,
4919 causing mysterious errors if for some reason the permissions on /dev/null
4920 got screwed. Now Exim writes a message to the main and panic logs, and
4921 bombs out if it can't open /dev/null.
4923 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4924 interact so that it is all more flexible. It is supposed to remain
4925 backwards compatible. Also added extra_local_interfaces.
4927 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4928 to bomb out with an assertion failure - to the client this appears as a
4929 connection drop. This problem occurs in the part of the code that was taken
4930 from the Samba project. Fortunately, the assertion is in a very simple
4931 function, so I have fixed this by reproducing the function inline in the
4932 one place where it is called, and arranging for authentication to fail
4933 instead of killing the process with assert().
4935 75. The SPA client code was not working when the server requested OEM rather
4936 than Unicode encoding.
4938 76. Added code to make require_files with a specific uid setting more usable in
4939 the case where statting the file as root fails - usually a non-root-mounted
4940 NFS file system. When this happens and the failure is EACCES, Exim now
4941 forks a subprocess and does the per-uid checking as the relevant uid.
4943 77. Added process_log_path.
4945 78. If log_file_path was not explicitly set, a setting of check_log_space or
4946 check_log_inodes was ignored.
4948 79. If a space check for the spool or log partitions fails, the incident is now
4949 logged. Of course, in the latter case the data may get lost...
4951 80. Added the %p formatting code to string_format() so that it can be used to
4952 print addresses in debug_print(). Adjusted all the address printing in the
4953 debugging in store.c to use %p rather than %d.
4955 81. There was a concern that a line of code in smtp_in.c could overflow a
4956 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4957 initially expressed, the concern was not well-founded, because trailing
4958 spaces are removed early. However, if the trailing spaces were followed by
4959 a NULL, they did not get removed, so the overflow was possible. Two fixes
4962 (a) I re-wrote the offending code in a cleaner fashion.
4963 (b) If an incoming SMTP command contains a NULL character, it is rejected
4966 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4967 runs initgroups(), so that if the Exim user is in any additional groups,
4968 they will be used during message reception.
4974 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob