1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.561 2009/01/02 17:12:03 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
98 TK/01 Add preliminary DKIM support. Currently requires a forked version of
99 ALT-N's libdkim that I have put here:
100 http://duncanthrax.net/exim-experimental/
102 Note to Michael Haardt: I had to rename some vars in sieve.c. They
103 were called 'true' and it seems that C99 defines that as a reserved
104 keyword to be used with 'bool' variable types. That means you could
105 not include C99-style headers which use bools without triggering
106 build errors in sieve.c.
108 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
109 as mailq or other aliases. Changed the --help handling significantly
110 to do whats expected. exim_usage() emits usage/help information.
112 SC/01 Added the -bylocaldomain option to eximstats.
114 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
116 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
118 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
124 PH/01 Another patch from the Sieve maintainer.
126 PH/02 When an IPv6 address is converted to a string for single-key lookup
127 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
128 dots are used instead of colons so that keys in lsearch files need not
129 contain colons. This was done some time before quoting was made available
130 in lsearch files. However, iplsearch files do require colons in IPv6 keys
131 (notated using the quote facility) so as to distinguish them from IPv4
132 keys. This meant that lookups for IP addresses in host lists did not work
133 for iplsearch lookups.
135 This has been fixed by arranging for IPv6 addresses to be expressed with
136 colons if the lookup type is iplsearch. This is not incompatible, because
137 previously such lookups could never work.
139 The situation is now rather anomolous, since one *can* have colons in
140 ordinary lsearch keys. However, making the change in all cases is
141 incompatible and would probably break a number of configurations.
143 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
146 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
147 conversion specification without a maximum field width, thereby enabling
148 a rogue spamd server to cause a buffer overflow. While nobody in their
149 right mind would setup Exim to query an untrusted spamd server, an
150 attacker that gains access to a server running spamd could potentially
151 exploit this vulnerability to run arbitrary code as the Exim user.
153 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
154 $primary_hostname instead of what libspf2 thinks the hosts name is.
156 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
157 a directory entry by the name of the lookup key. Previously, if a
158 symlink pointed to a non-existing file or a file in a directory that
159 Exim lacked permissions to read, a lookup for a key matching that
160 symlink would fail. Now it is enough that a matching directory entry
161 exists, symlink or not. (Bugzilla 503.)
163 PH/03 The body_linecount and body_zerocount variables are now exported in the
166 PH/04 Added the $dnslist_matched variable.
168 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
169 This means they are set thereafter only if the connection becomes
172 PH/06 Added the client_condition to authenticators so that some can be skipped
173 by clients under certain conditions.
175 PH/07 The error message for a badly-placed control=no_multiline_responses left
176 "_responses" off the end of the name.
178 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
180 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
181 (without spaces) instead of just copying the configuration text.
183 PH/10 Added the /noupdate option to the ratelimit ACL condition.
185 PH/11 Added $max_received_linelength.
187 PH/12 Added +ignore_defer and +include_defer to host lists.
189 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
190 way in which PCRE > 7.0 is built.
192 PH/14 Implemented queue_only_load_latch.
194 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
195 MAIL command. The effect was to mangle the value on 64-bit systems.
197 PH/16 Another patch from the Sieve maintainer.
199 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
201 PH/18 If a system quota error occurred while trying to create the file for
202 a maildir delivery, the message "Mailbox is full" was not appended to the
203 bounce if the delivery eventually timed out. Change 4.67/27 below applied
204 only to a quota excession during the actual writing of the file.
206 PH/19 It seems that peer DN values may contain newlines (and other non-printing
207 characters?) which causes problems in log lines. The DN values are now
208 passed through string_printing() before being added to log lines.
210 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
211 and InterBase are left for another time.)
213 PH/21 Added message_body_newlines option.
215 PH/22 Guard against possible overflow in moan_check_errorcopy().
217 PH/23 POSIX allows open() to be a macro; guard against that.
219 PH/24 If the recipient of an error message contained an @ in the local part
220 (suitably quoted, of course), incorrect values were put in $domain and
221 $local_part during the evaluation of errors_copy.
227 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
228 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
231 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
232 issue a MAIL command.
234 PH/02 In an ACL statement such as
236 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
238 if a client was not listed at all, or was listed with a value other than
239 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
240 the condition was not true (as it should be), so access was not denied.
241 The bug was that the ! inversion was incorrectly passed on to the second
242 item. This has been fixed.
244 PH/03 Added additional dnslists conditions == and =& which are different from
245 = and & when the dns lookup returns more than one IP address.
247 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
248 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
250 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
251 FSYNC, which compiles an option called disable_fsync that allows for
252 bypassing fsync(). The documentation is heavily laced with warnings.
254 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
256 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
257 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
258 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
259 including adding "make clean"; (3) Added -fPIC when compiling the test
260 dynamically loaded module, to get rid of a warning.
262 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
263 message fails, move_frozen_messages = true and ignore_bounce_errors_after
264 = 0s. The bug is otherwise harmless.
266 PH/07 There was a bug in the dovecot authenticator such that the value of
267 $auth1 could be overwritten, and so not correctly preserved, after a
268 successful authentication. This usually meant that the value preserved by
269 the server_setid option was incorrect.
271 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
273 PH/09 Installed PCRE release 7.0.
275 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
276 run for batched SMTP input. It is now run at the start of every message
277 in the batch. While fixing this I discovered that the process information
278 (output by running exiwhat) was not always getting set for -bs and -bS
279 input. This is fixed, and it now also says "batched" for BSMTP.
281 PH/11 Added control=no_pipelining.
283 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
284 patch, slightly modified), and move the expansion of helo_data till after
285 the connection is made in the smtp transport (so it can use these
288 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
290 PH/14 Added log_selector = +pid.
292 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
294 PH/16 Add ${if forany and ${if forall.
296 PH/17 Added dsn_from option to vary the From: line in DSNs.
298 PH/18 Flush SMTP output before performing a callout, unless control =
299 no_callout_flush is set.
301 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
302 was true (the default) a successful delivery failed to delete the retry
303 item, thus causing premature timeout of the address. The bug is now
306 PH/20 Added hosts_avoid_pipelining to the smtp transport.
308 PH/21 Long custom messages for fakedefer and fakereject are now split up
309 into multiline reponses in the same way that messages for "deny" and
310 other ACL rejections are.
312 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
313 with slight modification.
315 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
316 draft, changing the syntax and factoring some duplicate code".
318 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
319 for deliveries of the second and subsequent messages over the same SMTP
322 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
323 ${reduce, with only minor "tidies".
325 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
327 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
328 expansion side effects.
330 PH/27 When a message times out after an over-quota error from an Exim-imposed
331 quota, the bounce message says "mailbox is full". This message was not
332 being given when it was a system quota that was exceeded. It now should
335 MH/03 Made $recipients available in local_scan(). local_scan() already has
336 better access to the recipient list through recipients_list[], but
337 $recipients can be useful in postmaster-provided expansion strings.
339 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
340 in the case of a MAIL command with additional options following the
341 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
342 were accidentally chopped off.
344 PH/29 SMTP synchronization checks are implemented when a command is read -
345 there is a check that no more input is waiting when there shouldn't be
346 any. However, for some commands, a delay in an ACL can mean that it is
347 some time before the response is written. In this time, more input might
348 arrive, invalidly. So now there are extra checks after an ACL has run for
349 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
350 pipelining has not been advertised.
352 PH/30 MH's patch to allow iscntrl() characters to be list separators.
354 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
355 returned in the SMTP response when smtp_return_error_details was false.
358 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
359 instead of the C I/O that was originally supplied, because problems were
362 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
363 Exim which did not show up earlier: it was assuming that a call to
364 SSL_CTX_set_info_callback() might give an error value. In fact, there is
365 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
366 was a macro that became an assignment, so it seemed to work. This has
367 changed to a proper function call with a void return, hence the compile
368 error. Exim's code has been fixed.
370 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
373 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
375 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
377 PH/37 If a message is not accepted after it has had an id assigned (e.g.
378 because it turns out to be too big or there is a timeout) there is no
379 "Completed" line in the log. When some messages of this type were
380 selected by exigrep, they were listed as "not completed". Others were
381 picked up by some special patterns. I have improved the selection
382 criteria to be more general.
384 PH/38 The host_find_failed option in the manualroute router can now be set
385 to "ignore", to completely ignore a host whose IP address cannot be
386 found. If all hosts are ignored, the behaviour is controlled by the new
387 host_all_ignored option.
389 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
390 homing or because of multiple MX records with /mx) generated more than
391 one IP address, and the following item turned out to be the local host,
392 all the secondary addresses of the first item were incorrectly removed
393 from the list, along with the local host and any following hosts (which
394 is what is supposed to happen).
396 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
397 whoever called Exim into the -H file. In the case of the daemon it was
398 behaving confusingly. When first started, it used values for whoever
399 started the daemon, but after a SIGHUP it used the Exim user (because it
400 calls itself on a restart). I have changed the code so that it now always
403 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
404 message are rejected with the same error (e.g. no authentication or bad
405 sender address), and a DATA command is nevertheless sent (as can happen
406 with PIPELINING or a stupid MUA), the error message that was given to the
407 RCPT commands is included in the rejection of the DATA command. This is
408 intended to be helpful for MUAs that show only the final error to their
411 PH/42 Another patch from the Sieve maintainer.
413 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
414 Eximstats - Fixed some broken HTML links and added missing column headers
416 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
417 columns for Rejects, Temp Rejects, Ham, and Spam rows.
419 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
421 PH/43 Yet another patch from the Sieve maintainer.
423 PH/44 I found a way to check for a TCP/IP connection going away before sending
424 the response to the final '.' that terminates a message, but only in the
425 case where the client has not sent further data following the '.'
426 (unfortunately, this is allowed). However, in many cases there won't be
427 any further data because there won't be any more messages to send. A call
428 to select() can be used: if it shows that the input is "ready", there is
429 either input waiting, or the socket has been closed. An attempt to read
430 the next input character can distinguish the two cases. Previously, Exim
431 would have sent an OK response which the client would never have see.
432 This could lead to message repetition. This fix should cure that, at
433 least in a lot of common cases.
435 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
436 advertised in response to EHLO.
442 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
443 fixed by 4.65/MH/01 (is this a record?) are fixed:
445 (i) An empty string was always treated as zero by the numeric comparison
446 operators. This behaviour has been restored.
448 (ii) It is documented that the numeric comparison operators always treat
449 their arguments as decimal numbers. This was broken in that numbers
450 starting with 0 were being interpreted as octal.
452 While fixing these problems I realized that there was another issue that
453 hadn't been noticed. Values of message_size_limit (both the global option
454 and the transport option) were treated as octal if they started with 0.
455 The documentation was vague. These values are now always treated as
456 decimal, and I will make that clear in the documentation.
462 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
463 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
466 MH/01 Don't check that the operands of numeric comparison operators are
467 integers when their expansion is in "skipping" mode (fixes bug
468 introduced by 4.64-PH/07).
470 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
471 child addresses, Exim now panics and dies. Previously, because the count
472 is held in a short int, deliveries were likely to be lost. As such a
473 large number of recipients for a single message is ridiculous
474 (performance will be very, very poor), I have chosen to impose a limit
475 rather than extend the field.
481 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
482 leftover -K file (the existence of which was triggered by #402).
483 While we were at it, introduced process PID as part of the -K
484 filename. This should rule out race conditions when creating
487 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
488 processing considerably. Previous code took too long for large mails,
489 triggering a timeout which in turn triggers #401.
491 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
492 in the DK code in transports.c. sendfile() is not really portable,
493 hence the _LINUX specificness.
495 TF/01 In the add_headers option to the mail command in an Exim filter,
496 there was a bug that Exim would claim a syntax error in any
497 header after the first one which had an odd number of characters
500 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
501 callout verification, Exim cached a "reject" for the entire domain. This
502 is correct for most verifications, but it is not correct for a recipient
503 verification with use_sender or use_postmaster set, because in that case
504 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
505 case of MAIL FROM:<> rejection from other early rejections (e.g.
506 rejection of HELO). When verifying a recipient using a non-null MAIL
507 address, the cache is ignored if it shows MAIL FROM:<> rejection.
508 Whatever the result of the callout, the value of the domain cache is
509 left unchanged (for any other kind of callout, getting as far as trying
510 RCPT means that the domain itself is ok).
512 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
515 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
516 manifest itself as EPIPE rather than ECONNECT. When tidying away a
517 session, the daemon ignores ECONNECT errors and logs others; it now
518 ignores EPIPE as well.
520 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
521 (quoted-printable decoding).
523 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
524 later the small subsequent patch to fix an introduced bug.
526 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
528 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
530 PH/08 An error is now given if message_size_limit is specified negative.
532 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
533 to be given (somewhat) arbitrary names.
535 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
538 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
539 miscellaneous code fixes
541 PH/10 Added the log_reject_target ACL modifier to specify where to log
544 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
545 hostname. This is wrong, because it relates to the incoming message (and
546 probably the interface on which it is arriving) and not to the outgoing
547 callout (which could be using a different interface). This has been
548 changed to use the value of the helo_data option from the smtp transport
549 instead - this is what is used when a message is actually being sent. If
550 there is no remote transport (possible with a router that sets up host
551 addresses), $smtp_active_hostname is used.
553 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
554 tweaks were necessary in order to get it to work (see also 21 below):
555 (a) The code assumed that strncpy() returns a negative number on buffer
556 overflow, which isn't the case. Replaced with Exim's string_format()
558 (b) There were several signed/unsigned issues. I just did the minimum
559 hacking in of casts. There is scope for a larger refactoring.
560 (c) The code used strcasecmp() which is not a standard C function.
561 Replaced with Exim's strcmpic() function.
562 (d) The code set only $1; it now sets $auth1 as well.
563 (e) A simple test gave the error "authentication client didn't specify
564 service in request". It would seem that Dovecot has changed its
565 interface. Fortunately there's a specification; I followed it and
566 changed what the client sends and it appears to be working now.
568 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
571 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
572 address A is aliased to B and C, where B exists and C does not. Without
573 -v the output is "A verified" because verification stops after a
574 successful redirection if more than one address is generated. However,
575 with -v the child addresses are also verified. Exim was outputting "A
576 failed to verify" and then showing the successful verification for C,
577 with its parentage. It now outputs "B failed to verify", showing B's
578 parentage before showing the successful verification of C.
580 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
581 look up a TXT record in a specific list after matching in a combined
584 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
585 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
586 they consult the DNS. I had assumed they would set it the way they
587 wanted; and indeed my experiments on Linux seem to show that in some
588 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
589 To be on the safe side, however, I have now made the interface to
590 host_find_byname() similar to host_find_bydns(), with an argument
591 containing the DNS resolver options. The host_find_byname() function now
592 sets these options at its start, just as host_find_bydns() does. The smtp
593 transport options dns_qualify_single and dns_search_parents are passed to
594 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
595 of host_find_byname() use the default settings of RES_DEFNAMES
596 (qualify_single) but not RES_DNSRCH (search_parents).
598 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
599 spool_read_header() do less string testing, by means of a preliminary
600 switch on the second character of optional "-foo" lines. (This is
601 overdue, caused by the large number of possibilities that now exist.
602 Originally there were few.) While I was there, I also converted the
603 str(n)cmp tests so they don't re-test the leading "-" and the first
604 character, in the hope this might squeeze out yet more improvement.
606 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
607 flag allowing group syntax was set by the header_syntax check but not
608 turned off, possible causing trouble later; (2) The flag was not being
609 set at all for the header_verify test, causing "group"-style headers to
610 be rejected. I have now set it in this case, and also caused header_
611 verify to ignore an empty address taken from a group. While doing this, I
612 came across some other cases where the code for allowing group syntax
613 while scanning a header line wasn't quite right (mostly, not resetting
614 the flag correctly in the right place). These bugs could have caused
615 trouble for malformed header lines. I hope it is now all correct.
617 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
618 with the "reply" argument non-NULL. The code, however (which originally
619 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
620 but it didn't always do it. This confused somebody who was copying the
621 code for some other use. I have removed all the tests.
623 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
624 feature that was used to support insecure browsers during the U.S. crypto
625 embargo. It requires special client support, and Exim is probably the
626 only MTA that supported it -- and would never use it because real RSA is
627 always available. This code has been removed, because it had the bad
628 effect of slowing Exim down by computing (never used) parameters for the
629 RSA_EXPORT functionality.
631 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
632 authenticator to fail if there's a tab character in the incoming data
633 (there should never be unless someone is messing about, as it's supposed
634 to be base64-encoded). Also added, on Timo's advice, the "secured" option
635 if the connection is using TLS or if the remote IP is the same as the
636 local IP, and the "valid-client-cert option" if a client certificate has
639 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
640 authenticators. This can be used for authorization after authentication
641 succeeds. (In the case of plaintext, it servers for both authentication
644 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
645 if any retry times were supplied.
647 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
648 connection, where there is no client IP address to check. In this
649 situation, the verify now always succeeds.
651 PH/25 Applied John Jetmore's -Mset patch.
653 PH/26 Added -bem to be like -Mset, but loading a message from a file.
655 PH/27 In a string expansion for a processed (not raw) header when multiple
656 headers of the same name were present, leading whitespace was being
657 removed from all of them, but trailing whitespace was being removed only
658 from the last one. Now trailing whitespace is removed from each header
659 before concatenation. Completely empty headers in a concatenation (as
662 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
663 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
665 PH/29 [Removed. This was a change that I later backed out, and forgot to
666 correct the ChangeLog entry (that I had efficiently created) before
667 committing the later change.]
669 PH/30 Exim was sometimes attempting to deliver messages that had suffered
670 address errors (4xx response to RCPT) over the same connection as other
671 messages routed to the same hosts. Such deliveries are always "forced",
672 so retry times are not inspected. This resulted in far too many retries
673 for the affected addresses. The effect occurred only when there were more
674 hosts than the hosts_max_try setting in the smtp transport when it had
675 the 4xx errors. Those hosts that it had tried were not added to the list
676 of hosts for which the message was waiting, so if all were tried, there
677 was no problem. Two fixes have been applied:
679 (i) If there are any address or message errors in an SMTP delivery, none
680 of the hosts (tried or untried) are now added to the list of hosts
681 for which the message is waiting, so the message should not be a
682 candidate for sending over the same connection that was used for a
683 successful delivery of some other message. This seems entirely
684 reasonable: after all the message is NOT "waiting for some host".
685 This is so "obvious" that I'm not sure why it wasn't done
686 previously. Hope I haven't missed anything, but it can't do any
687 harm, as the worst effect is to miss an optimization.
689 (ii) If, despite (i), such a delivery is accidentally attempted, the
690 routing retry time is respected, so at least it doesn't keep
691 hammering the server.
693 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
694 in ${readsocket because some servers need this prod.
696 PH/32 Added some extra debug output when updating a wait-xxx database.
698 PH/33 The hint "could be header name not terminated by colon", which has been
699 given for certain expansion errors for a long time, was not being given
700 for the ${if def:h_colon_omitted{... case.
702 PH/34 The spec says: "With one important exception, whenever a domain list is
703 being scanned, $domain contains the subject domain." There was at least
704 one case where this was not true.
706 PH/35 The error "getsockname() failed: connection reset by peer" was being
707 written to the panic log as well as the main log, but it isn't really
708 panic-worthy as it just means the connection died rather early on. I have
709 removed the panic log writing for the ECONNRESET error when getsockname()
712 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
713 runs only) independently of the message's sender address. This meant
714 that, if the 4xx error was in fact related to the sender, a different
715 message to the same recipient with a different sender could confuse
716 things. In particualar, this can happen when sending to a greylisting
717 server, but other circumstances could also provoke similar problems.
718 I have changed the default so that the retry time for these errors is now
719 based a combination of the sender and recipient addresses. This change
720 can be overridden by setting address_retry_include_sender=false in the
723 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
724 remote server are returned as part of bounce messages. This was not
725 happening for LMTP over a pipe (the lmtp transport), but now it is the
726 same for both kinds of LMTP.
728 PH/38 Despite being documented as not happening, Exim was rewriting addresses
729 in header lines that were in fact CNAMEs. This is no longer the case.
731 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
732 and queue runs started by the daemon processed all messages. This has
733 been fixed so that -R and -S can now usefully be given with -q<time>.
735 PH/40 Import PCRE release 6.7 (fixes some bugs).
737 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
739 PH/42 Give an error if -q is specified more than once.
741 PH/43 Renamed the variables $interface_address and $interface_port as
742 $received_ip_address and $received_port, to make it clear that these
743 values apply to message reception, and not to the outgoing interface when
744 a message is delivered. (The old names remain recognized, of course.)
746 PH/44 There was no timeout on the connect() call when using a Unix domain
747 socket in the ${readsocket expansion. There now is.
749 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
750 be meaningful with "accept".
752 SC/01 Eximstats V1.43
753 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
755 SC/02 Eximstats V1.44
756 Use a glob alias rather than an array ref in the generated
757 parser. This improves both readability and performance.
759 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
760 Collect SpamAssassin and rejection statistics.
761 Don't display local sender or destination tables unless
762 there is data to show.
763 Added average volumes into the top table text output.
765 SC/04 Eximstats V1.46
766 Collect data on the number of addresses (recipients)
767 as well as the number of messages.
769 SC/05 Eximstats V1.47
770 Added 'Message too big' to the list of mail rejection
771 reasons (thanks to Marco Gaiarin).
773 SC/06 Eximstats V1.48
774 Mainlog lines which have GMT offsets and are too short to
775 have a flag are now skipped.
777 SC/07 Eximstats V1.49 (Alain Williams)
778 Added the -emptyok flag.
780 SC/08 Eximstats V1.50
781 Fixes for obtaining the IP address from reject messages.
783 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
784 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
785 whitesspace changes from 4.64-PH/27
787 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
790 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
791 are found, allow negative numbers in numeric criteria)
793 JJ/06 exipick.20061117.2, added new $message_body_missing variable
795 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
796 to match changes made in 4.64-PH/43
798 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
800 PH/47 Put in an explicit test for a DNS lookup of an address record where the
801 "domain" is actually an IP address, and force a failure. This locks out
802 those revolvers/nameservers that support "A-for-A" lookups, in
803 contravention of the specifications.
805 PH/48 When a host name was looked up from an IP address, and the subsequent
806 forward lookup of the name timed out, the host name was left in
807 $sender_host_name, contrary to the specification.
809 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
810 restricted to single-key lookups, Exim was not diagnosing an error if
811 * or *@ was used with a query-style lookup.
813 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
815 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
816 long ago, but noone interested enough thought of it. Let's just say that
817 the "1.1" means that there are some new functions that weren't there at
818 some point in the past.
820 PH/51 Error processing for expansion failure of helo_data from an smtp
821 transport during callout processing was broken.
823 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
824 tested/used via the -bh/-bhc/-bs options.
826 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
827 bug, fixed in subsequent PCRE releases).
829 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
830 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
832 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
838 SC/01 Use a glob alias rather than an array ref in eximstats generated
839 parser. This improves both readability and performance.
841 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
842 Don't display local sender or destination tables in eximstats unless
843 there is data to show.
844 Added average volumes into the eximstats top table text output.
846 SC/03 Collect data on the number of addresses (recipients) as well
847 as the number of messages in eximstats.
849 TF/01 Correct an error in the documentation for the redirect router. Exim
850 does (usually) call initgroups() when daemonizing.
852 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
853 with consistent privilege compared to when running as a daemon.
855 TF/03 Note in the spec that $authenticated_id is not set for local
856 submissions from trusted users.
858 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
859 Thanks to Dean Brooks <dean@iglou.com> for the patch.
861 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
862 by adding some example configuration directives to the default
863 configuration file. A little bit of work is required to uncomment the
864 directives and define how usernames and passwords are checked, but
865 there is now a framework to start from.
867 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
868 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
869 without this. I don't know how relevant this is to other LDAP libraries.
871 PH/02 Add the verb name to the "unknown ACL verb" error.
873 PH/03 Magnus Holmgren's patch for filter_prepend_home.
875 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
877 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
878 directory not expanded when it should be if an expanded home directory
879 was set for the address (which is overridden by the transport).
881 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
884 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
885 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
886 because it is too late at that time, and has no effect.
888 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
889 security issue with \' (bugzilla #107). I could not use the
890 PQescapeStringConn() function, because it needs a PGconn value as one of
893 PH/08 When testing addresses using -bt, indicate those final addresses that
894 are duplicates that would not cause an additional delivery. At least one
895 person was confused, thinking that -bt output corresponded to deliveries.
896 (Suppressing duplicates isn't a good idea as you lose the information
897 about possibly different redirections that led to the duplicates.)
899 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
900 systems where poll() doesn't work, in particular OS X.
902 PH/10 Added more information to debugging output for retry time not reached.
904 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
905 operations in malware.c.
907 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
910 PH/13 If write_rejectlog was set false when logging was sent to syslog with
911 syslog_duplication set false, log lines that would normally be written
912 both the the main log and to the reject log were not written to syslog at
915 PH/14 In the default configuration, change the use of "message" in ACL warn
916 statements to "add_header".
918 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
919 not followed by a command (e.g. "seen endif").
921 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
922 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
925 PH/17 Added extra conditions to the default value of delay_warning_condition
929 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
930 { match{$h_precedence:}{(?i)bulk|list|junk} } \
931 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
934 The Auto-Submitted: and various List- headers are standardised, whereas I
935 don't think Precedence: ever was.
937 PH/18 Refactored debugging code in route_finduser() to show more information,
938 in particular, the error code if getpwnam() issues one.
940 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
941 This is apparently needed in addition to the PH/07 change above to avoid
942 any possible encoding problems.
944 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
945 but not after initializing Perl.
947 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
948 output them only if debugging. By default they are written stderr,
949 apparently, which is not desirable.
951 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
954 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
957 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
959 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
960 authenticated or an ident call has been made. Suppress the default
961 values for $authenticated_id and $authenticated_sender (but permit -oMai
962 and -oMas) when testing with -bh.
964 PH/24 Re-jigged the order of the tests in the default configuration so that the
965 tests for valid domains and recipients precede the DNS black list and CSA
966 tests, on the grounds that those ones are more expensive.
968 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
969 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
970 This bug exists in every version of Exim that I still have, right back to
973 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
974 However, an attempt to turn on case-sensitivity in a regex key by
975 including (?-i) didn't work because the subject string was already
976 lowercased, and the effects were non-intuitive. It turns out that a
977 one-line patch can be used to allow (?-i) to work as expected.
983 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
984 other effects) broke the use of negated acl sub-conditions.
986 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
989 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
990 "Deny" causes Exim to reject the incoming connection with a 554 error.
991 Unfortunately, if there is a major crisis, such as a disk failure,
992 tcp-wrappers gives "deny", whereas what one would like would be some
993 kind of temporary error. A kludge has been added to help with this.
994 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
995 554 error is used if errno is still zero or contains ENOENT (which occurs
996 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
999 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1001 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1002 errors. Otherwise a message that provokes a temporary error (when other
1003 messages do not) can cause a whole host to time out.
1005 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1006 addresses were routed directly to files or pipes from a redirect router.
1007 File deliveries just didn't batch; pipe deliveries might have suffered
1010 PH/06 A failure to get a lock for a hints database would erroneously always say
1011 "Failed to get write lock", even when it was really a read lock.
1013 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1014 of 0600. This has been changed to use the value of the lockfile_mode
1015 option (which defaults to 0600).
1017 PH/08 Applied small patch from the Sieve maintainer.
1019 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1020 folder from quota calculations, a direct delivery into this folder messed
1021 up the contents of the maildirsize file. This was because the regex was
1022 used only to exclude .Trash (or whatever) when the size of the mailbox
1023 was calculated. There was no check that a delivery was happening into an
1024 excluded directory. This bug has been fixed by ignoring all quota
1025 processing for deliveries into excluded directories.
1027 PH/10 Added the maildirfolder_create_regex option to appendfile.
1033 PH/01 The code for finding all the local interface addresses on a FreeBSD
1034 system running IPv6 was broken. This may well have applied to all BSD
1035 systems, as well as to others that have similar system calls. The broken
1036 code found IPv4 interfaces correctly, but gave incorrect values for the
1037 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1038 that it would not match correctly against @[] and not recognize the IPv6
1041 PH/02 The ipliteral router was not recognizing addresses of the form user@
1042 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1044 PH/03 Added disable_ipv6.
1046 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1047 decoded form, because it is most often used to construct To: headers
1048 lines in autoreplies, and the decoded form may well be syntactically
1049 invalid. However, $reply_address has leading white space removed, and all
1050 newlines turned into spaces so that the autoreply transport does not
1053 PH/05 If group was specified without a user on a router, and no group or user
1054 was specified on a transport, the group from the router was ignored.
1056 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1057 for visible compile-time settings that can be used to change these
1058 numbers, for those that want even more. Backwards compatibility with old
1059 spool files has been maintained. However, going back to a previous Exim
1060 release will lost any variables that are in spool files.
1062 PH/07 Two small changes when running in the test harness: increase delay when
1063 passing a TCP/IP connection to a new process, in case the original
1064 process has to generate a bounce, and remove special handling of
1065 127.0.0.2 (sic), which is no longer necessary.
1067 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1068 be the same on different OS.
1070 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1073 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1074 whether --show-vars was specified or not
1076 JJ/02 exipick: Added support for new ACL variable spool format introduced
1079 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1080 syntactically invalid From: or Reply-to: line, and a filter used this to
1081 generate an autoreply, and therefore failed to obtain an address for the
1082 autoreply, Exim could try to deliver to a non-existent relative file
1083 name, causing unrelated and misleading errors. What now happens is that
1084 it logs this as a hard delivery error, but does not attempt to create a
1087 PH/11 The exinext utility has a -C option for testing purposes, but although
1088 the given file was scanned by exinext itself; it wasn't being passed on
1089 when Exim was called.
1091 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1092 an end-of-file indication when reading a command response.
1094 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1095 compiled. In many other places in Exim, IPv6 addresses are always
1096 recognized, so I have changed this. It also means that IPv4 domain
1097 literals of the form [IPV4:n.n.n.n] are now always recognized.
1099 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1100 used if the router is not running as root, for example, when verifying at
1101 ACL time, or when using -bh. The debugging output from this situation was
1102 non-existent - all you got was a failure to exec. I have made two
1105 (a) Failures to set uid/gid, the current directory, or a process leader
1106 in a subprocess such as that created by queryprogram now generate
1107 suitable debugging ouput when -d is set.
1109 (b) The queryprogram router detects when it is not running as root,
1110 outputs suitable debugging information if -d is set, and then runs
1111 the subprocess without attempting to change uid/gid.
1113 PH/15 Minor change to Makefile for building test_host (undocumented testing
1116 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1117 additional section of a DNS packet that returns MX or SRV records.
1118 Instead, it always explicitly searches for A/AAAA records. This avoids
1119 major problems that occur when a DNS server includes only records of one
1120 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1121 fixed another bug: if SRV records were looked up and the corresponding
1122 address records were *not* found in the additional section, the port
1123 values from the SRV records were lost.
1125 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1126 using the correct key (the original address) when searching the retry
1127 rules in order to find which one to use for generating the retry hint.
1129 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1130 adding the default one. Similarly, if it contains a Reply-To: header, the
1131 errors_reply_to option, if set, is not used.
1133 PH/19 When calculating a retry time, Exim used to measure the "time since
1134 failure" by looking at the "first failed" field in the retry record. Now
1135 it does not use this if it is later than than the arrival time of the
1136 message. Instead it uses the arrival time. This makes for better
1137 behaviour in cases where some deliveries succeed, thus re-setting the
1138 "first failed" field. An example is a quota failure for a huge message
1139 when small messages continue to be delivered. Without this change, the
1140 "time since failure" will always be short, possible causing more frequent
1141 delivery attempts for the huge message than are intended.
1142 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1144 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1145 $1, $2, $3) because the numerical variables can be reset during some
1146 expansion items (e.g. "match"), thereby losing the authentication data.
1148 PH/21 Make -bV show the size of off_t variables so that the test suite can
1149 decide whether to run tests for quotas > 2G.
1151 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1152 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1153 filecount value is greater than 2G or if a quota value is greater than 2G
1154 on a system where the size of off_t is not greater than 4, a panic error
1157 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1158 never match. The debug and -bh output now contains an explicit error
1159 message indicating a malformed IPv4 address or mask.
1161 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1162 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1163 PH/23 above applies.
1165 PH/25 Do not write to syslog when running in the test harness. The only
1166 occasion when this arises is a failure to open the main or panic logs
1167 (for which there is an explicit test).
1169 PH/26 Added the /no_tell option to "control=freeze".
1171 PH/27 If a host name lookup failed very early in a connection, for example, if
1172 the IP address matched host_lookup and the reverse lookup yielded a name
1173 that did not have a forward lookup, an error message of the form "no IP
1174 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1175 could be logged. Now it outputs the IP address instead of "NULL".
1177 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1178 allows the sender and the authenticated sender to be set when
1179 submitting a message from within Exim. Since child_open_exim() is
1180 documented for local_scan(), the new function should be too.
1182 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1183 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1184 results in an empty string is now treated as unset.
1186 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1188 PH/31 Added sender_verify_fail logging option.
1190 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1191 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1192 tidied the source and removed it altogether.
1194 PH/33 When a queue run was abandoned because the load average was too high, a
1195 log line was always written; now it is written only if the queue_run log
1196 selector is set. In addition, the log line for abandonment now contains
1197 information about the queue run such as the pid. This is always present
1198 in "start" and "stop" lines but was omitted from the "abandon" line.
1200 PH/34 Omit spaces between a header name and the colon in the error message that
1201 is given when verify = headers_syntax fails (if there are lots of them,
1202 the message gets confusing).
1204 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1205 names, as there are now some PTR records that contain slashes. This check
1206 is only to protect against broken name servers that fall over on strange
1207 characters, so the fact that it applies to all lookups doesn't matter.
1209 PH/36 Now that the new test suite is complete, we can remove some of the
1210 special code in Exim that was needed for the old test suite. For example,
1211 sorting DNS records because real resolvers return them in an arbitrary
1212 order. The new test suite's fake resolver always returns records in the
1215 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1216 bounces) except when queue_only is set, to avoid logging races between
1217 the different processes.
1219 PH/38 Panic-die if .include specifies a non-absolute path.
1221 PH/39 A tweak to the "H" retry rule from its user.
1223 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1224 a label. They prevented compilation on older perls.
1226 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1227 a warning to be raised on newish perls.
1229 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1230 on queue. Changes to match documented behaviour of showing count of
1231 messages matching specified criteria.
1233 PH/40 Changed the default ident timeout from 30s to 5s.
1235 PH/41 Added support for the use of login_cap features, on those BSD systems
1236 that have them, for controlling the resources used by pipe deliveries.
1238 PH/42 The content-scanning code uses fopen() to create files in which to put
1239 message data. Previously it was not paying any attention to the mode of
1240 the files. Exim runs with umask(0) because the rest of the code creates
1241 files with open(), and sets the required mode explicitly. Thus, these
1242 files were ending up world-writeable. This was not a big issue, because,
1243 being within the spool directory, they were not world-accessible. I have
1244 created a function called modefopen, which takes an additional mode
1245 argument. It sets umask(777), creates the file, chmods it to the required
1246 mode, then resets the umask. All the relevant calls to fopen() in the
1247 content scanning code have been changed to use this function.
1249 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1250 to 24 hours. This avoids potential overflow problems when processing G
1251 and H retry rules. I suspect nobody ever tinkers with this value.
1253 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1255 PH/45 When the plaintext authenticator is running as a client, the server's
1256 challenges are checked to ensure they are valid base64 strings. By
1257 default, the authentication attempt is cancelled if an invalid string is
1258 received. Setting client_ignore_invalid_base64 true ignores these errors.
1259 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1260 they are received. Thus, the responses can be made to depend on the
1261 challenges. If an invalid string is ignored, an empty string is placed in
1264 PH/46 Messages that are created by the autoreply transport now contains a
1265 References: header, in accordance with RFCs 2822 and 3834.
1267 PH/47 Added authenticated_sender_force to the smtp transport.
1269 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1271 PH/49 Installed latest patch from the Sieve maintainer.
1273 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1274 was also set, the appendfile transport was unnecessarily scanning a
1275 directory of message files (e.g. for maildir delivery) to find the count
1276 of files (along with the size), even though it did not need this
1277 information. It now does the scan only if it needs to find either the
1278 size of the count of files.
1280 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1282 PH/52 Two bugs concerned with error handling when the smtp transport is
1285 (i) Exim was not creating retry information for temporary errors given
1286 for individual recipients after the DATA command when the smtp transport
1287 was used in LMTP mode. This meant that they could be retried too
1288 frequently, and not timed out correctly.
1290 (ii) Exim was setting the flag that allows error details to be returned
1291 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1292 recipients that were returned after the DATA command.
1294 PH/53 This is related to PH/52, but is more general: for any failing address,
1295 when detailed error information was permitted to be returned to the
1296 sender, but the error was temporary, then after the final timeout, only
1297 "retry timeout exceeded" was returned. Now it returns the full error as
1298 well as "retry timeout exceeded".
1300 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1301 do this, and (what is worse) MTAs that accept it.
1303 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1304 will now be deprecated.
1306 PH/56 New os.c-cygwin from the Cygwin maintainer.
1308 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1309 formats (previously only available in exim formats via -bpr, -bpru,
1310 and -bpra. Now also available in native and exiqgrep formats)
1312 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1313 with very large, slow to parse queues
1315 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1317 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1319 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1320 responses to authentication challenges, though it was showing the
1321 challenges; (ii) I've removed the CR characters from the debug output for
1324 PH/58 Allow for the insertion of a newline as well as a space when a string
1325 is turned into more than one encoded-word during RFC 2047 encoding. The
1326 Sieve code now uses this.
1328 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1329 data_4xx, lost_connection, tls_required.
1331 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1332 message was being sent as an SMTP response.
1334 PH/61 Add -l and -k options to exicyclog.
1336 PH/62 When verifying, if an address was redirected to one new address, so that
1337 verification continued, and the new address failed or deferred after
1338 having set something in $address_data, the value of $address_data was not
1339 passed back to the ACL. This was different to the case when no
1340 redirection occurred. The value is now passed back in both cases.
1342 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1343 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1344 use login_cap.h, so on its own it isn't the distinguishing feature. The
1345 new name refers directly to the setclassresources() function.
1347 PH/65 Added configuration files for NetBSD3.
1349 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1351 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1352 is preferred over IPv4.
1354 PH/68 The bounce_return_message and bounce_return_body options were not being
1355 honoured for bounces generated during the reception of non-SMTP messages.
1356 In particular, this applied to messages rejected by the ACL. This bug has
1357 been fixed. However, if bounce_return_message is true and bounce_return_
1358 body is false, the headers that are returned for a non-SMTP message
1359 include only those that have been read before the error was detected.
1360 (In the case of an ACL rejection, they have all been read.)
1362 PH/69 The HTML version of the specification is now built in a directory called
1363 spec_html instead of spec.html, because the latter looks like a path with
1364 a MIME-type, and this confuses some software.
1366 PH/70 Catch two compiler warnings in sieve.c.
1368 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1369 function verify_get_ident() calls ip_connect() to connect a socket, but
1370 if the "connect()" function timed out, ip_connect() used to close the
1371 socket. However, verify_get_ident() also closes the socket later, and in
1372 between Exim writes to the log, which may get opened at this point. When
1373 the socket was closed in ip_connect(), the log could get the same file
1374 descriptor number as the socket. This naturally causes chaos. The fix is
1375 not to close the socket in ip_connect(); the socket should be closed by
1376 the function that creates it. There was only one place in the code where
1377 this was missing, in the iplookup router, which I don't think anybody now
1378 uses, but I've fixed it anyway.
1380 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1381 well as to direct DNS lookups. Otherwise the handling of names in host
1382 lists is inconsistent and therefore confusing.
1388 PH/01 Two changes to the default runtime configuration:
1390 (1) Move the checks for relay_from_hosts and authenticated clients from
1391 after to before the (commented out) DNS black list checks.
1393 (2) Add control=submission to the relay_from_hosts and authenticated
1394 clients checks, on the grounds that messages accepted by these
1395 statements are most likely to be submissions.
1397 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1399 (1) Generate an error if the third argument for the ${prvs expansion is
1402 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1405 (3) Reset the variables that are obtained from the first argument of
1406 ${prvscheck and used in the second argument before leaving the code,
1407 because their memory is reclaimed, so using them afterwards may do
1410 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1411 one (it's much easier than Tom thought :-).
1413 (5) Because of (4), we can now allow for the use of $prvscheck_result
1414 inside the third argument.
1416 PH/03 For some reason, the default setting of PATH when running a command from
1417 a pipe transport was just "/usr/bin". I have changed it to
1420 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1421 anything to be listed in the output from -bV.
1423 PH/05 When a filter generated an autoreply, the entire To: header line was
1424 quoted in the delivery log line, like this:
1426 => >A.N.Other <ano@some.domain> <original@ddress> ...
1428 This has been changed so that it extracts the operative address. There
1429 may be more than one such address. If so, they are comma-separated, like
1432 => >ano@some.domain,ona@other.domain <original@ddress> ...
1434 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1435 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1436 not being looked up in the rDNS to get a host name, Exim was showing the
1437 IP address twice in Received: lines, even though the IP addresses were
1438 identical. For example:
1440 Received: from [1.2.3.4] (helo=[1.2.3.4])
1442 However, if the real host name was known, it was omitting the HELO data
1443 if it matched the actual IP address. This has been tidied up so that it
1444 doesn't show the same IP address twice.
1446 PH/07 When both +timestamp and +memory debugging was on, the value given by
1447 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1448 called by the debug printing, thereby overwriting the timestamp buffer.
1449 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1451 PH/08 When the original message was included in an autoreply transport, it
1452 always said "this is a copy of the message, including all the headers",
1453 even if body_only or headers_only was set. It now gives an appropriate
1456 PH/09 Applied a patch from the Sieve maintainer which:
1458 o fixes some comments
1459 o adds the (disabled) notify extension core
1460 o adds some debug output for the result of if/elsif tests
1461 o points to the current vacation draft in the documentation
1462 and documents the missing references header update
1466 o fixes a bug in processing the envelope test (when testing
1467 multiple envelope elements, the last element determinted the
1470 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1471 Electronic Mail") by including:
1473 Auto-submitted: auto-generated
1475 in the messages that it generates (bounce messages and others, such as
1476 warnings). In the case of bounce messages for non-SMTP mesages, there was
1477 also a typo: it was using "Auto_submitted" (underscore instead of
1478 hyphen). Since every message generated by Exim is necessarily in response
1479 to another message, thes have all been changed to:
1481 Auto-Submitted: auto-replied
1483 in accordance with these statements in the RFC:
1485 The auto-replied keyword:
1487 - SHOULD be used on messages sent in direct response to another
1488 message by an automatic process,
1490 - MUST NOT be used on manually-generated messages,
1492 - MAY be used on Delivery Status Notifications (DSNs) and Message
1493 Disposition Notifications (MDNs),
1495 - MUST NOT be used on messages generated by automatic or periodic
1496 processes, except for messages which are automatic responses to
1499 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1500 to the default Received: header definition.
1502 PH/12 Added log selector acl_warn_skipped (default on).
1504 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1505 variables because (a) they are in the wrong storage pool and (b) even if
1506 they were copied, it wouldn't work properly because of the caching.
1508 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1509 checking when decoding. Apparently there are clients that generate
1510 overlong encoded strings. Why am I not surprised?
1512 PH/15 If the first argument of "${if match_address" was not empty, but did not
1513 contain an "@" character, Exim crashed. Now it writes a panic log message
1514 and treats the condition as false.
1516 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1518 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1519 extension "envelope-auth". The code is finished and in agreement with
1520 other implementations, but there is no documentation so far and in fact,
1521 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1522 not changing the active code.
1524 Print executed "if" and "elsif" statements when debugging is used. This
1525 helps a great deal to understand what a filter does.
1527 Document more things not specified clearly in RFC3028. I had all this
1528 sorted out, when out of a sudden new issues came to my mind. Oops."
1530 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1533 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1534 canonical form (as documented). However, after a host name lookup from
1535 the IP address, check_host() was doing a simple string comparison with
1536 addresses acquired from the DNS when checking that the found name did
1537 have the original IP as one of its addresses. Since any found IPv6
1538 addresses are likely to be in abbreviated form, the comparison could
1539 fail. Luckily, there already exists a function for doing the comparison
1540 by converting both addresses to binary, so now that is used instead of
1541 the text comparison.
1543 PH/20 There was another similar case to PH/19, when a complete host name was
1544 given in a host list; looking up its IP address could give an abbreviated
1545 form, whereas the current host's name might or might not be abbreviated.
1546 The same fix has been applied.
1552 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1553 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1556 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1557 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1559 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1561 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1562 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1563 Also, for "Auto-Submitted", treat anything other than "no" as
1564 non-personal, in accordance with RFC 3834. (Previously it treated
1565 anything starting "auto-" as non-personal.)
1567 TF/01 The control=submission/name=... option had a problem with syntax
1568 errors if the name included a slash character. The /name= option
1569 now slurps the rest of the string, so it can include any characters
1570 but it must come last in the list of options (after /sender_retain
1573 PH/05 Some modifications to the interface to the fake nameserver for the new
1581 TK/01 Added the "success_on_redirect" address verification option. See
1582 NewStuff for rationale and an example.
1584 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1586 PH/02 Patch to exigrep to allow it to work on syslog lines.
1588 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1589 fread() to skip over the body file's header line, because in Cygwin the
1590 header line is locked and is inaccessible.
1592 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1593 co-exist for some time) to make it clear that it is the Exim ID that is
1594 referenced, not the Message-ID: header line.
1596 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1597 string_format() function, because snprintf() does not exist on all
1600 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1601 ${stat: expansion item.
1603 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1604 protocol synchronization error", to keep the pedants happy.
1606 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1607 well as for IRIX systems, when gcc is being used. See the host.c source
1610 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1612 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1615 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1616 transport and to the smtp transport in LMTP mode.
1618 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1620 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1621 run a filter in a subprocess. This could lead to confusion in subsequent
1622 lookups in the parent process. There should also be a search_tidyup() at
1623 the end of the subprocess.
1625 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1626 only if the host matched helo_try_verify_hosts, which caused the
1627 verification to occur when the EHLO/HELO command was issued. The ACL just
1628 tested the remembered result. Now, if a previous verification attempt has
1629 not happened, "verify = helo" does it there and then.
1631 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1633 TK/03 Fix log output including CR from clamd.
1635 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1636 exist provoked a memory error which could cause a segfault.
1638 PH/15 Installed PCRE 6.2
1640 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1642 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1643 of the problem. Specifically, suggested +O2 rather than +O1 for the
1646 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1648 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1649 record, the port was not correctly logged when the outgoing_port log
1650 selector was set (it logged the transort's default port).
1652 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1653 fallback_hosts, and "hosts" in the smtp transport.
1655 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1656 host errors such as "Connection refused".
1658 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1659 authentication with radiusclient 0.4.9:
1661 - Error returned from rc_read_config was caught wrongly
1662 - Username/password not passed on to radius server due to wrong length.
1664 The presumption is that some radiusclient API changes for 4.51/PH/17
1665 were not taken care of correctly. The code is still untested by me (my
1666 Linux distribution still has 0.3.2 of radiusclient), but it was
1667 contributed by a Radius user.
1669 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1670 expanding the "port" option of the smtp transport.
1672 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1673 while reading a MIME header. Thanks to Tom Hughes for a patch.
1675 PH/24 Include config.h inside local_scan.h so that configuration settings are
1678 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1679 that in an ACL for RCPT (for example), you can examine exactly what was
1682 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1683 commands, but it was not correctly comparing the address with the actual
1684 client host address. Thus, it would show the EHLO address in Received:
1685 header lines when this was not necessary.
1687 PH/27 Added the % operator to ${eval:}.
1689 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1690 it should be ignoring failures (because with -C, for example, it has lost
1691 privilege). It wasn't ignoring creation failures other than "already
1694 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1697 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1698 because an input line was too long, either on its own, or by virtue of
1699 too many continuations, the temporary file was not being removed, and the
1700 return code was incorrect.
1702 PH/31 Missing "BOOL" in function definition in filtertest.c.
1704 PH/32 Applied Sieve patches from the maintainer.
1706 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1708 PH/33 Added "verify = not_blind".
1710 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1711 Local/Makefile (with some defaults set). These are used in built scripts
1712 such as exicyclog, but they have never been used in the exim_install
1713 script (though there are many overriding facilities there). I have
1714 arranged that the exim_install script now takes note of these two
1717 PH/35 Installed configuration files for Dragonfly.
1719 PH/36 When a locally submitted message by a trusted user did not contain a
1720 From: header, and the sender address was obtained from -f or from an SMTP
1721 MAIL command, and the trusted user did not use -F to supply a sender
1722 name, $originator_name was incorrectly used when constructing a From:
1723 header. Furthermore, $originator_name was used for submission mode
1724 messages from external hosts without From: headers in a similar way,
1725 which is clearly wrong.
1727 PH/37 Added control=suppress_local_fixups.
1729 PH/38 When log_selector = +received_sender was set, and the addition of the
1730 sender made the log line's construction buffer exactly full, or one byte
1731 less than full, an overflow happened when the terminating "\n" was
1734 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1735 when the result of a list match is failure because a DNS lookup failed.
1737 PH/40 RM_COMMAND is now used in the building process.
1739 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1740 the "build-* directories that it finds.
1742 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1743 address in a domain literal was a prefix of an interface address.
1745 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1746 when verifying a sender address, unless rewrite_headers is false.
1748 PH/44 Wrote a long comment about why errors_to addresses are verified as
1749 recipients, not senders.
1751 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1752 the ratelimit ACL was added.
1754 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1756 PH/46 Added extra information about PostgreSQL errors to the error string.
1758 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1759 avoiding the need to install special zones in a real server. This is
1760 backwards compatible; if it can't find the fake resolver, it drops back.
1761 Thus, both old and new test suites can be run.
1763 TF/02 Added util/ratelimit.pl
1765 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1766 clock is set back in time.
1768 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1769 Candler <B.Candler@pobox.com>.
1771 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1772 OK for addresses that are the result of redirections.
1774 PH/48 A number of further additions for the benefit of the new test suite,
1775 including a fake gethostbyname() that interfaces to the fake DNS resolver
1778 TF/06 The fix for widen_domains has also been applied to qualify_single and
1779 search_parents which are the other dnslookup options that can cause
1782 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1785 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1787 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1788 service" if it encountered an MX record with an empty target hostname.
1789 The message is now "an MX or SRV record indicated no SMTP service".
1791 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1792 if the DNS of the sending site is misconfigured. This is quite a
1793 common situation. This change restores the behaviour of treating a
1794 helo verification defer as a failure.
1796 PH/51 If self=fail was set on a router, the bounce message did not include the
1797 actual error message.
1803 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1805 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1806 command itself ended in error, the underlying message about the transport
1807 filter timeout was being overwritten with the pipe command error. Now the
1808 underlying error message should be appended to the second error message.
1810 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1812 PH/02 Reduce the amount of output that "make" produces by default. Full output
1813 can still be requested.
1815 PH/03 The warning log line about a condition test deferring for a "warn" verb
1816 was being output only once per connection, rather than after each
1817 occurrence (because it was using the same function as for successful
1818 "warn" verbs). This seems wrong, so I have changed it.
1820 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1821 it should not have, which might have caused a crash in the right
1822 circumstances, but probably never did.
1824 PH/04 Installed a modified version of Tony Finch's patch to make submission
1825 mode fix the return path as well as the Sender: header line, and to
1826 add a /name= option so that you can make the user's friendly name appear
1829 TF/03 Added the control = fakedefer ACL modifier.
1831 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1832 Mark Lowes for thorough testing.
1834 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1836 TK/03 Merged latest SRS patch from Miles Wilton.
1838 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1839 with the definition in sysexits.h (which is #included earlier).
1840 Fortunately, Exim does not actually use EX_OK. The code used to try to
1841 preserve the sysexits.h value, by assumimg that macro definitions were
1842 scanned for macro replacements. I have been disabused of this notion,
1843 so now the code just undefines EX_OK before #including unistd.h.
1845 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1846 in the smtp transport. When a block could not be written in a single
1847 write() function, the timeout was being re-applied to each part-write.
1848 This seems wrong - if the receiver was accepting one byte at a time it
1849 would take for ever. The timeout is now adjusted when this happens. It
1850 doesn't have to be particularly precise.
1852 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1853 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1855 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1856 without a domain if the check to <postmaster@domain> fails.
1858 SC/01 Eximstats: added -xls and the ability to specify output files
1859 (patch written by Frank Heydlauf).
1861 SC/02 Eximstats: use FileHandles for outputing results.
1863 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1865 SC/04 Eximstats: fixed display of large numbers with -nvr option
1867 SC/05 Eximstats: fixed merging of reports with empty tables.
1869 SC/06 Eximstats: added the -include_original_destination flag
1871 SC/07 Eximstats: removed tabs and trailing whitespace.
1873 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1875 TK/06 MBOX spool code: Add real "From " MBOX separator line
1876 so the .eml file is really in mbox format (even though
1877 most programs do not really care). Patch from Alex Miller.
1879 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1880 The latter is generated from $received_to and is only set if the
1881 message has one envelope recipient. SA can use these headers,
1882 obviously out-of-the-box. Patch from Alex Miller.
1884 PH/08 The ${def test on a variable was returning false if the variable's
1885 value was "0", contrary to what the specification has always said!
1886 The result should be true unless the variable is empty.
1888 PH/09 The syntax error of a character other than { following "${if
1889 def:variable_name" (after optional whitespace) was not being diagnosed.
1890 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1891 accidental colon was present, for example, could give incorrect results.
1893 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1894 result is used (not including appendfile, where other changes are about
1897 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1898 This involved changing a lot of size variables from int to off_t. It
1899 should work with maildirs and everything.
1901 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1902 spamd dying while we are connected to it.
1904 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1907 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1908 function for BDB 4.3.
1910 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1912 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1913 the sources for PCRE have been re-arranged and more files are now
1916 PH/15 The code I had for printing potentially long long variables in PH/11
1917 above was not the best (it lost precision). The length of off_t variables
1918 is now inspected at build time, and an appropriate printing format (%ld
1919 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1920 to be "long long int" or "long int". This is needed for the internal
1921 formatting function string_vformat().
1923 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1924 the configuration file to be ":syslog", then the script "guesses" where
1925 the logs files are, rather than using the compiled in default. In our
1926 case the guess is not the same as the compiled default, so the script
1927 suddenly stopped working when I started to use syslog. The patch checks
1928 to see if log_file_path is "". If so, it attempts to read it from exim
1929 with no configuration file to get the compiled in version, before it
1930 falls back to the previous guessing code."
1932 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1933 implementing BATV in an Exim configuration. See NewStuff for the gory
1936 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1937 Makefile that are specific to HP-UX.
1939 PH/18 If the "use_postmaster" option was set for a recipient callout together
1940 with the "random" option, the postmaster address was used as the MAIL
1941 FROM address for the random test, but not for the subsequent recipient
1942 test. It is now used for both.
1944 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1945 patch removes a few documentation additions to RFC 3028, because the
1946 latest draft now contains them. It adds the new en;ascii-case comparator
1947 and a new error check for 8bit text in MIME parts. Comparator and
1948 require names are now matched exactly. I enabled the subaddress
1949 extension, but it is not well tested yet (read: it works for me)."
1951 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1952 rework some of the code of TK/09 above to avoid the hardwired use of
1953 "%lld" and "long long". Replaced the call to snprintf() with a call to
1956 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1957 records point to non-existent hosts", "retry timeout exceeded", and
1958 "retry time not reached for any host after a long failure period".
1960 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1961 experimental DomainKeys support:
1963 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1964 (2) On an error such as an illegally used "control", the wrong name for
1965 the control was given.
1967 These problems did NOT occur unless DomainKeys support was compiled.
1969 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1971 PH/24 Added ${if match_ip condition.
1973 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1974 databases so that it will be absolutely obvious if a crash occurs in the
1975 DB library. This is a regular occurrence (often caused by mis-matched
1978 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1979 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1980 picked up on a user's system that detects such things. There doesn't seem
1981 to be a gcc warning option for this - only an attribute that has to be
1982 put on the function's prototype. It seems that in Fedora Core 4 they have
1983 set this on a number of new functions. No doubt there will be more in due
1986 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1987 specify a transport. However, if an address that was verified by such a
1988 router was the subject of a callout, Exim crashed because it tried to
1989 read the rcpt_include_affixes from the non-existent transport. Now it
1990 just assumes that the setting of that option is false. This bug was
1991 introduced by 4.51/PH/31.
1993 PH/28 Changed -d+all to exclude +memory, because that information is very
1994 rarely of interest, but it makes the output a lot bigger. People tend to
1995 do -d+all out of habit.
1997 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1998 code in os-type was giving problems when libc.so lives in lib64, like on
2001 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2002 aren't the modern standard, and it seems that some systems' include files
2003 don't always have them. Exim was already checking for some of the newer
2004 ones like T_AAAA, and defining it itself. I've added checks for all the
2005 record types that Exim uses.
2007 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2008 not automatically generating a new one, as it is supposed to. This
2009 prevented TLS from working. If the file did exist, but contained invalid
2010 data, a new version was generated, as expected. It was only the case of a
2011 non-existent file that was broken.
2013 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2014 with a change in libdomainkeys > 0.64.
2016 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2017 from DNS. If the selector record carries the flag, it now has
2018 precedence over the domain-wide flag.
2020 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2022 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2023 the use of a port name that isn't defined in /etc/services) occurred, the
2024 message was deferred as in a normal delivery, and thus remained on the
2025 spool, instead of being failed because of the mua_wrapper setting. This
2026 is now fixed, and I tidied up some of the mua_wrapper messages at the
2029 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2030 the messages in a hash of arrays rather than using individual hashes.
2031 This is a bit cleaner and results in dramatic memory savings, albeit
2032 at a slight CPU cost.
2034 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2035 as requested by Marc Sherman.
2037 SC/10 Eximstats: added histograms for user specified patterns as requested
2040 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2042 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2043 fopen() in the content-scanning modules that did not already have it.
2049 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2050 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2052 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2054 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2056 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2057 to test Sieve filters that use "vacation".
2059 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2060 that changes the way the GnuTLS parameters are stored in the cache file.
2061 The new format can be generated externally. For backward compatibility,
2062 if the data in the cache doesn't make sense, Exim assumes it has read an
2063 old-format file, and it generates new data and writes a new file. This
2064 means that you can't go back to an older release without removing the
2067 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2068 work if there are any delivery delays because "one_time" forces the
2069 parent to be marked "delivered", so its unseen clone is never tried
2070 again. For this reason, Exim now forbids the simultaneous setting of
2073 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2074 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2075 entry if you want to know the details. The fix, however, neglected to
2076 consider the case when local delivery batching is involved. The test for
2077 "previously delivered" was not happening when checking to see if an
2078 address could be batched with a previous (undelivered) one; under
2079 certain circumstances this could lead to multiple deliveries to the same
2082 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2083 in its include files, and this causes problems building Exim.
2085 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2086 header_syntax) but Exim was just ignoring anything given after a slash.
2087 In particular, this caused confusion with an attempt to use "verify =
2088 reverse_host_lookup/defer_ok". An error is now given when options are
2089 supplied for verify items that do not have them. (Maybe reverse_host_
2090 lookup should have a defer_ok option, but that's a different point.)
2092 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2093 defined by RFC 821) to 2048, because there were problems with some AUTH
2094 commands, and RFC 1869 says the size should be increased for extended
2095 SMTP commands that take arguments.
2097 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2100 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2101 "unknown" error; now it says that the functionality isn't in the binary.
2103 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2104 an address' error message when a string expansion fails (syntax or
2105 whatever). Otherwise the password may appear in the log. Following change
2106 PH/42 below, there is no longer a chance of it appearing in a bounce
2109 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2111 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2112 of its addresses was ever tried. (Bugzilla bug #2.)
2114 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2115 the result incorrectly in the debug output. (It correctly added a newline
2116 to what was transported.)
2118 TF/01 Added $received_time.
2120 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2121 commented out examples of how to interface to a virus scanner and to
2122 SpamAssassin. Also added commented examples of av_scanner and
2123 spamd_address settings.
2125 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2126 and controls are allowed in which ACLs. There were a couple of minor
2127 errors. Some of the entries in the conditions table (which is a table of
2128 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2129 negation of where the condition IS allowed.
2131 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2133 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2134 header file does not have a version number, so I've had to invent a new
2135 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2136 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2137 radiusclient), but it was contributed by a Radius user.
2139 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2140 files or directories, for OpenSSL.
2142 PH/19 When an Exim process that is running as root has to create an Exim log
2143 file, it does so in a subprocess that runs as exim:exim so as to get the
2144 ownership right at creation (otherwise, other Exim processes might see
2145 the file with the wrong ownership). There was no test for failure of this
2146 fork() call, which would lead to the process getting stuck as it waited
2147 for a non-existent subprocess. Forks do occasionally fail when resources
2148 run out. I reviewed all the other calls to fork(); they all seem to check
2151 PH/20 When checking for unexpected SMTP input at connect time (before writing
2152 the banner), Exim was not dealing correctly with a non-positive return
2153 from the read() function. If the client had disconnected by this time,
2154 the result was a log entry for a synchronization error with an empty
2155 string after "input=" when read() returned zero. If read() returned -1
2156 (an event I could not check), uninitialized data bytes were printed.
2157 There were reports of junk text (parts of files, etc) appearing after
2160 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2162 PH/22 Added support for macro redefinition, and (re)definition in between
2163 driver and ACL definitions.
2165 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2166 forgetting to use the resulting value; it was using the unexpanded value.
2168 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2169 hadn't been configured. The fix is from Juergen Kreileder, who
2170 understands it better than I do:
2172 "Here's what I see happening with three configured cyrus_sasl
2173 authenticators configured (plain, login, cram-md5):
2175 On startup auth_cyrus_sasl_init() gets called for each of these.
2176 This means three calls to sasl_listmech() without a specified mech_list.
2177 => SASL tests which mechs of all available mechs actually work
2178 => three warnings about OTP not working
2179 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2181 With the patch, sasl_listmech() also gets called three times. But now
2182 SASL's mech_list option is set to the server_mech specified in the the
2183 authenticator. Or in other words, the answer from sasl_listmech()
2184 gets limited to just the mech you're testing for (which is different
2186 => the return list contains just 'plain' or 'login', 'cram-md5' or
2187 nothing depending on the value of ob->server_mech.
2189 I've just tested the patch: Authentication still works fine,
2190 unavailable mechs specified in the exim configuration are still
2191 caught, and the auth.log warnings about OTP are gone."
2193 PH/25 When debugging is enabled, the contents of the command line are added
2194 to the debugging output, even when log_selector=+arguments is not
2197 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2198 answer is "GNU", and only if the return is "GNU/something" is the answer
2201 PH/27 $acl_verify_message is now set immediately after the failure of a
2202 verification in an ACL, and so is available in subsequent modifiers. In
2203 particular, the message can be preserved by coding like this:
2205 warn !verify = sender
2206 set acl_m0 = $acl_verify_message
2208 Previously, $acl_verify_message was set only while expanding "message"
2209 and "log_message" when a very denied access.
2211 PH/28 Modified OS/os.c-Linux with
2213 -#ifndef OS_LOAD_AVERAGE
2214 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2216 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2217 nomenclature these days.)
2219 PH/29 Installed patch from the Sieve maintainer that adds the options
2220 sieve_useraddress and sieve_subaddress to the redirect router.
2222 PH/30 In these circumstances:
2223 . Two addresses routed to the same list of hosts;
2224 . First host does not offer TLS;
2225 . First host accepts first address;
2226 . First host gives temporary error to second address;
2227 . Second host offers TLS and a TLS session is established;
2228 . Second host accepts second address.
2229 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2230 and peerdn, if requested) that were in fact used only for the second
2233 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2234 attention to any local part prefix or suffix that was matched by the
2235 router that accepted the address. It now behaves in the same way as it
2236 does for delivery: the affixes are removed from the local part unless
2237 rcpt_include_affixes is set on the transport.
2239 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2240 timeout during the DATA phase of an incoming message.
2242 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2243 applied a patch sanctioned by the Sieve maintainer.
2245 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2246 the uid or gid is negative. A case of a negative gid caused this to be
2247 noticed. The fix allows for either to be negative.
2249 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2250 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2253 PH/36 The change PH/12 above was broken. Fixed it.
2255 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2256 the grounds that routing the same address twice would always produce the
2257 same answer. This might have been true once, but it is certainly no
2258 longer true now. Routing a child address may depend on the previous
2259 routing that produced that child. Some complicated redirection strategies
2260 went wrong when messages had multiple recipients, and made Exim's
2261 behaviour dependent on the order in which the addresses were given.
2263 I have moved the duplicate checking until after the routing is complete.
2264 Exim scans the addresses that are assigned to local and remote
2265 transports, and removes any duplicates. This means that more work will be
2266 done, as duplicates will always all be routed, but duplicates are
2267 presumably rare, so I don't expect this is of any significance.
2269 For deliveries to pipes, files, and autoreplies, the duplicate checking
2270 still happens during the routing process, since they are not going to be
2273 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2274 It corrects a timeout issue with spamd. This is Ian's comment: "The
2275 background is that sometimes spamd either never reads data from a
2276 connection it has accepted, or it never writes response data. The exiscan
2277 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2278 blindly assumes that writes won't block so it may never time out."
2280 PH/39 Allow G after quota size as well as K and M.
2282 PH/40 The value set for $authenticated_id in an authenticator may not contain
2283 binary zeroes or newlines because the value is written to log lines and
2284 to spool files. There was no check on this. Now the value is run through
2285 the string_printing() function so that such characters are converted to
2286 printable escape sequences.
2288 PH/41 $message_linecount is a new variable that contains the total number of
2289 lines in the message. Compare $body_linecount, which is the count for the
2292 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2293 bounce and delay warning messages, except in certain special cases, which
2296 (a) An SMTP error message from a remote host;
2297 (b) A message specified in a :fail: redirection;
2298 (c) A message specified in a "fail" command in a system filter;
2299 (d) A message specified in a FAIL return from the queryprogram router;
2300 (e) A message specified by the cannot_route_message router option.
2302 In these cases only, Exim does include the error details in bounce and
2303 warning messages. There are also a few cases where bland messages such
2304 as "unrouteable address" or "local delivery error" are given.
2306 PH/43 $value is now also set for the "else" part of a ${run expansion.
2308 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2309 being worked on, but at least Exim now implements the latest version to
2312 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2313 process to complete was treated as a delivery failure, a timeout while
2314 writing the message to the pipe was logged, but erroneously treated as a
2315 successful delivery. Such timeouts include transport filter timeouts. For
2316 consistency with the overall process timeout, these timeouts are now
2317 treated as errors, giving rise to delivery failures by default. However,
2318 there is now a new Boolean option for the pipe transport called
2319 timeout_defer, which, if set TRUE, converts the failures into defers for
2320 both kinds of timeout. A transport filter timeout is now identified in
2323 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2324 systems where "make" and "gmake" are different, calling "gmake" at top
2325 level broke things. I've arranged for the value of $(MAKE) to be passed
2326 from the Makefile to this script so that it can call the same version of
2330 A note about Exim versions 4.44 and 4.50
2331 ----------------------------------------
2333 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2334 changes of various kinds. As a consequence, a big documentation update was
2335 needed. This delayed the release for rather longer than seemed good, especially
2336 in the light of a couple of (minor) security issues. Therefore, the changes
2337 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2338 release. So 4.44 and 4.50 are in effect two different branches that both start
2341 I have left the 4.50 change log unchanged; it contains all the changes since
2342 4.43. The change log for 4.44 is below; many of its items are identical to
2343 those for 4.50. This seems to be the most sensible way to preserve the
2344 historical information.
2350 1. Minor wording change to the doc/README.SIEVE file.
2352 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2353 computation of the current number of files was incorrect.
2355 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2356 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2359 4. Give more explanation in the error message when the command for a transport
2360 filter fails to execute.
2362 5. There are several places where Exim runs a non-Exim command in a
2363 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2364 was being done only for the command run by the queryprogram router. It is
2365 now done for all such subprocesses. The other cases are: ${run, transport
2366 filters, and the commands run by the lmtp and pipe transports.
2368 6. Added CONFIGURE_GROUP build-time option.
2370 7. Some older OS have a limit of 256 on the maximum number of file
2371 descriptors. Exim was using setrlimit() to set 1000 as a large value
2372 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2373 systems. I've change it so that if it can't get 1000, it tries for 256.
2375 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2376 was an oversight, and furthermore, ever since the addition of extra
2377 controls (e.g. 4.43/32), the checks on when to allow different forms of
2378 "control" were broken. There should now be diagnostics for all cases when a
2379 control that does not make sense is encountered.
2381 9. Added the /retain_sender option to "control=submission".
2383 10. $recipients is now available in the predata ACL (oversight).
2385 11. Tidy the search cache before the fork to do a delivery from a message
2386 received from the command line. Otherwise the child will trigger a lookup
2387 failure and thereby defer the delivery if it tries to use (for example) a
2388 cached ldap connection that the parent has called unbind on.
2390 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2391 of $address_data from the recipient verification was clobbered by the
2392 sender verification.
2394 13. The value of address_data from a sender verification is now available in
2395 $sender_address_data in subsequent conditions in the ACL statement.
2397 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2399 15. Added a new option "connect=<time>" to callout options, to set a different
2402 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2403 was its contents. (It was OK if the option was not defined at all.)
2405 17. A "Completed" log line is now written for messages that are removed from
2406 the spool by the -Mrm option.
2408 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2409 information about exactly what failed.
2411 19. Added -dd to debug only the daemon process.
2413 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2414 handles timeouts, both on the server side and network timeouts. Renamed the
2415 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2417 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2418 It is now set to "smtps".
2420 22. $host_address is now set to the target address during the checking of
2421 ignore_target_hosts.
2423 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2424 being passed; this would have caused $sender_host_name to have been used if
2425 matching the list had actually called for a host name (not very likely,
2426 since this list is usually IP addresses). A host name is now passed as
2429 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2430 code that specifies a non-restarting handler (typically sigaction() in
2431 modern systems) in an attempt to fix a rare and obscure crash bug.
2433 25. Narrowed the window for a race in the daemon that could cause it to ignore
2434 SIGCHLD signals. This is not a major problem, because they are used only to
2435 wake it up if nothing else does.
2437 26. A malformed maildirsize file could cause Exim to calculate negative values
2438 for the mailbox size or file count. Odd effects could occur as a result.
2439 The maildirsize information is now recalculated if the size or filecount
2442 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2443 support for a long time. Removed HAVE_SYS_VFS_H.
2445 28. Installed the latest version of exipick from John Jetmore.
2447 29. In an address list, if the pattern was not a regular expression, an empty
2448 subject address (from a bounce message) matched only if the pattern was an
2449 empty string. Non-empty patterns were not even tested. This was the wrong
2450 because it is perfectly reasonable to use an empty address as part of a
2451 database query. An empty address is now tested by patterns that are
2452 lookups. However, all the other forms of pattern expect the subject to
2453 contain a local part and a domain, and therefore, for them, an empty
2454 address still always fails if the pattern is not itself empty.
2456 30. Exim went into a mad DNS loop when attempting to do a callout where the
2457 host was specified on an smtp transport, and looking it up yielded more
2458 than one IP address.
2460 31. Re-factored the code for checking spool and log partition space into a
2461 function that finds that data and another that does the check. The former
2462 is then used to implement four new variables: $spool_space, $log_space,
2463 $spool_inodes, and $log_inodes.
2465 32. The RFC2047 encoding function was originally intended for short strings
2466 such as real names; it was not keeping to the 75-character limit for
2467 encoded words that the RFC imposes. It now respects the limit, and
2468 generates multiple encoded words if necessary. To be on the safe side, I
2469 have increased the buffer size for the ${rfc2047: expansion operator from
2472 33. It is now permitted to omit both strings after an "if" condition; if the
2473 condition is true, the result is "true". As before, when the second string
2474 is omitted, a false condition yields an empty string. This makes it less
2475 cumbersome to write custom ACL and router conditions.
2477 34. Failure to deliver a bounce message always caused it to be frozen, even if
2478 there was an errors_to setting on the router. The errors_to setting is now
2481 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2482 canonical form (fully expanded) before being placed in
2483 $sender_host_address.
2485 36. The table in the code that translates DNS record types into text (T_A to
2486 "A" for instance) was missing entries for NS and CNAME. It is just possible
2487 that this could have caused confusion if both these types were looked up
2488 for the same domain, because the text type is used as part of Exim's
2489 per-process caching. But the chance of anyone hitting this buglet seems
2492 37. The dnsdb lookup has been extended in a number of ways.
2494 (1) There is a new type, "zns", which walks up the domain tree until it
2495 finds some nameserver records. It should be used with care.
2497 (2) There is a new type, "mxh", which is like "mx" except that it returns
2498 just the host names, not the priorities.
2500 (3) It is now possible to give a list of domains (or IP addresses) to be
2501 looked up. The behaviour when one of the lookups defers can be
2502 controlled by a keyword.
2504 (4) It is now possible to specify the separator character for use when
2505 multiple records are returned.
2507 38. The dnslists ACL condition has been extended: it is now possible to supply
2508 a list of IP addresses and/or domains to be looked up in a particular DNS
2511 39. Added log_selector=+queue_time_overall.
2513 40. When running the queue in the test harness, wait just a tad after forking a
2514 delivery process, to get repeatability of debugging output.
2516 41. Include certificate and key file names in error message when GnuTLS fails
2517 to set them up, because the GnuTLS error message doesn't include the name
2518 of the failing file when there is a problem reading it.
2520 42. Allow both -bf and -bF in the same test run.
2522 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2524 44. The "Exiscan patch" is now merged into the mainline Exim source.
2526 45. Sometimes the final signoff response after QUIT could fail to get
2527 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2528 < 0 before doing a fflush(). This bug looks as though it goes back to the
2529 introduction of TLS in release 3.20, but "sometimes" must have been rare
2530 because the tests only now provoked it.
2532 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2533 (this can affect the format of dates).
2535 47. exim_tidydb, when checking for the continued existence of a message for
2536 which it has found a message-specific retry record, was not finding
2537 messages that were in split spool directories. Consequently, it was
2538 deleting retry records that should have stayed in existence.
2540 48. Steve fixed some bugs in eximstats.
2542 49. The SPA authentication driver was not abandoning authentication and moving
2543 on to the next authenticator when an expansion was forced to fail,
2544 contradicting the general specification for all authenticators. Instead it
2545 was generating a temporary error. It now behaves as specified.
2547 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2548 (the order specifies the preference for clients). The order is now AES256,
2549 AES128, 3DES, ARCFOUR128.
2551 51. Small patch to Sieve code - explicitly set From: when generating an
2554 52. Exim crashed if a remote delivery caused a very long error message to be
2555 recorded - for instance if somebody sent an entire SpamAssassin report back
2556 as a large number of 550 error lines. This bug was coincidentally fixed by
2557 increasing the size of one of Exim's internal buffers (big_buffer) that
2558 happened as part of the Exiscan merge. However, to be on the safe side, I
2559 have made the code more robust (and fixed the comments that describe what
2562 53. Now that there can be additional text after "Completed" in log lines (if
2563 the queue_time_overall log selector is set), a one-byte patch to exigrep
2564 was needed to allow it to recognize "Completed" as not the last thing in
2567 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2568 patch that reportedly fixes this has been added. I am not expert enough to
2569 create a test for it. This is what the patch creator wrote:
2571 "I found a little strange behaviour of ldap code when working with
2572 Windows 2003 AD Domain, where users was placed in more than one
2573 Organization Units. When I tried to give exim partial DN, the exit code
2574 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2575 But simultaneously result of request was absolutely normal ldap result,
2576 so I produce this patch..."
2578 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2579 REFERENCE, so I have modified the code to exclude the patch when that macro
2582 55. Some experimental protocols are using DNS PTR records for new purposes. The
2583 keys for these records are domain names, not reversed IP addresses. The
2584 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2585 leaves it alone. Component reversal etc. now happens only for IP addresses.
2588 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2590 57. Double the size of the debug message buffer (to 2048) so that more of very
2591 long debug lines gets shown.
2593 58. The exicyclog utility now does better if the number of log files to keep
2594 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2596 59. Two changes related to the smtp_active_hostname option:
2598 (1) $smtp_active_hostname is now available as a variable.
2599 (2) The default for smtp_banner uses $smtp_active_hostname instead
2600 of $primary_hostname.
2602 60. The host_aton() function is supposed to be passed a string that is known
2603 to be a valid IP address. However, in the case of IPv6 addresses, it was
2604 not checking this. This is a hostage to fortune. Exim now panics and dies
2605 if the condition is not met. A case was found where this could be provoked
2606 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2607 components; fortuitously, this particular loophole had already been fixed
2608 by change 4.50/55 above.
2610 If there are any other similar loopholes, the new check in host_aton()
2611 itself should stop them being exploited. The report I received stated that
2612 data on the command line could provoke the exploit when Exim was running as
2613 exim, but did not say which command line option was involved. All I could
2614 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2615 running as the user.
2618 61. There was a buffer overflow vulnerability in the SPA authentication code
2619 (which came originally from the Samba project). I have added a test to the
2620 spa_base64_to_bits() function which I hope fixes it.
2623 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2624 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2626 63. The daemon start-up calls getloadavg() while still root for those OS that
2627 need the first call to be done as root, but it missed one case: when
2628 deliver_queue_load_max is set with deliver_drop_privilege. This is
2629 necessary for the benefit of the queue runner, because there is no re-exec
2630 when deliver_drop_privilege is set.
2632 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2633 This has been fixed.
2635 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2636 was in use, was not putting the data itself into the right store pool;
2637 consequently, it could be overwritten for a subsequent message in the same
2638 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2641 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2643 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2644 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2645 the function were treating the return as a boolean value, which happened to
2646 work because 0=false and not-0=true, but is not correct code.
2648 68. The host_aton() function was not handling scoped IPv6 addresses (those
2649 with, for example, "%eth0" on the end) correctly.
2651 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2652 negated items (that is, ~something) in unsigned ints. Some compilers
2653 apparently mutter when there is no cast.
2655 70. If an address verification called from an ACL failed, and did not produce a
2656 user-specific message (i.e. there was only a "system" message), nothing was
2657 put in $acl_verify_message. In this situation, it now puts the system
2660 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2661 session; change 4.31/43 added the unwanted input to the log line - except
2662 that it did not do this in the start of session case. It now does.
2664 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2665 This is wrong and can cause the other end to generate a synchronization
2666 error if it is another Exim or anything else that does the synchronization
2667 check. A QUIT command is no longer sent after a timeout.
2669 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2670 during host lookups.
2672 74. The defer_ok option of callout verification was not working if it was used
2673 when verifying addresses in header lines, that is, for this case:
2675 verify = header_sender/callout=defer_ok
2677 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2678 those file descriptors could be used for SMTP connections. If anything
2679 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2680 could be sent to the SMTP client, causing chaos. The daemon now opens
2681 stdin, stdout, and stderr to /dev/null when it puts itself into the
2684 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2685 log by default. The user can override this with suitable Perl magic.
2687 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2688 the log message when discard triggers, was not working for the DATA ACL or
2689 for the non-SMTP ACL.
2691 78. Error message wording change in sieve.c.
2693 79. If smtp_accept_max_per_host was set, the number of connections could be
2694 restricted to fewer than expected, because the daemon was trying to set up
2695 a new connection before checking whether the processes handling previous
2696 connections had finished. The check for completed processes is now done
2697 earlier. On busy systems, this bug wouldn't be noticed because something
2698 else would have woken the daemon, and it would have reaped the completed
2701 80. If a message was submitted locally by a user whose login name contained one
2702 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2703 It caused a spool format error. I have fixed the spool reading code. A
2704 related problem was that the "from" clause in the Received: line became
2705 illegal because of the space(s). It is now covered by ${quote_local_part.
2707 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2710 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2712 83. Patch from Sieve maintainer to fix unterminated string problem in
2713 "vacation" handling.
2715 84. Some minor changes to the Linux configuration files to help with other
2716 OS variants using glibc.
2718 85. One more patch for Sieve to update vacation handling to latest spec.
2721 ----------------------------------------------------
2722 See the note above about the 4.44 and 4.50 releases.
2723 ----------------------------------------------------
2729 1. Change 4.43/35 introduced a bug that caused file counts to be
2730 incorrectly computed when quota_filecount was set in an appendfile
2733 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2734 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2737 3. Give more explanation in the error message when the command for a transport
2738 filter fails to execute.
2740 4. There are several places where Exim runs a non-Exim command in a
2741 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2742 was being done only for the command run by the queryprogram router. It is
2743 now done for all such subprocesses. The other cases are: ${run, transport
2744 filters, and the commands run by the lmtp and pipe transports.
2746 5. Some older OS have a limit of 256 on the maximum number of file
2747 descriptors. Exim was using setrlimit() to set 1000 as a large value
2748 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2749 systems. I've change it so that if it can't get 1000, it tries for 256.
2751 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2752 was an oversight, and furthermore, ever since the addition of extra
2753 controls (e.g. 4.43/32), the checks on when to allow different forms of
2754 "control" were broken. There should now be diagnostics for all cases when a
2755 control that does not make sense is encountered.
2757 7. $recipients is now available in the predata ACL (oversight).
2759 8. Tidy the search cache before the fork to do a delivery from a message
2760 received from the command line. Otherwise the child will trigger a lookup
2761 failure and thereby defer the delivery if it tries to use (for example) a
2762 cached ldap connection that the parent has called unbind on.
2764 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2765 of $address_data from the recipient verification was clobbered by the
2766 sender verification.
2768 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2769 was its contents. (It was OK if the option was not defined at all.)
2771 11. A "Completed" log line is now written for messages that are removed from
2772 the spool by the -Mrm option.
2774 12. $host_address is now set to the target address during the checking of
2775 ignore_target_hosts.
2777 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2778 being passed; this would have caused $sender_host_name to have been used if
2779 matching the list had actually called for a host name (not very likely,
2780 since this list is usually IP addresses). A host name is now passed as
2783 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2784 code that specifies a non-restarting handler (typically sigaction() in
2785 modern systems) in an attempt to fix a rare and obscure crash bug.
2787 15. Narrowed the window for a race in the daemon that could cause it to ignore
2788 SIGCHLD signals. This is not a major problem, because they are used only to
2789 wake it up if nothing else does.
2791 16. A malformed maildirsize file could cause Exim to calculate negative values
2792 for the mailbox size or file count. Odd effects could occur as a result.
2793 The maildirsize information is now recalculated if the size or filecount
2796 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2797 support for a long time. Removed HAVE_SYS_VFS_H.
2799 18. Updated exipick to current release from John Jetmore.
2801 19. Allow an empty sender to be matched against a lookup in an address list.
2802 Previously the only cases considered were a regular expression, or an
2805 20. Exim went into a mad DNS lookup loop when doing a callout where the
2806 host was specified on the transport, if the DNS lookup yielded more than
2809 21. The RFC2047 encoding function was originally intended for short strings
2810 such as real names; it was not keeping to the 75-character limit for
2811 encoded words that the RFC imposes. It now respects the limit, and
2812 generates multiple encoded words if necessary. To be on the safe side, I
2813 have increased the buffer size for the ${rfc2047: expansion operator from
2816 22. Failure to deliver a bounce message always caused it to be frozen, even if
2817 there was an errors_to setting on the router. The errors_to setting is now
2820 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2821 canonical form (fully expanded) before being placed in
2822 $sender_host_address.
2824 24. Updated eximstats to version 1.33
2826 25. Include certificate and key file names in error message when GnuTLS fails
2827 to set them up, because the GnuTLS error message doesn't include the name
2828 of the failing file when there is a problem reading it.
2830 26. Expand error message when OpenSSL has problems setting up cert/key files.
2833 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2834 (this can affect the format of dates).
2836 28. exim_tidydb, when checking for the continued existence of a message for
2837 which it has found a message-specific retry record, was not finding
2838 messages that were in split spool directories. Consequently, it was
2839 deleting retry records that should have stayed in existence.
2841 29. eximstats updated to version 1.35
2842 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2843 1.35 - bugfix such that pie charts by volume are generated correctly
2845 30. The SPA authentication driver was not abandoning authentication and moving
2846 on to the next authenticator when an expansion was forced to fail,
2847 contradicting the general specification for all authenticators. Instead it
2848 was generating a temporary error. It now behaves as specified.
2850 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2851 (the order specifies the preference for clients). The order is now AES256,
2852 AES128, 3DES, ARCFOUR128.
2854 31. Small patch to Sieve code - explicitly set From: when generating an
2857 32. Exim crashed if a remote delivery caused a very long error message to be
2858 recorded - for instance if somebody sent an entire SpamAssassin report back
2859 as a large number of 550 error lines. This bug was coincidentally fixed by
2860 increasing the size of one of Exim's internal buffers (big_buffer) that
2861 happened as part of the Exiscan merge. However, to be on the safe side, I
2862 have made the code more robust (and fixed the comments that describe what
2865 33. Some experimental protocols are using DNS PTR records for new purposes. The
2866 keys for these records are domain names, not reversed IP addresses. The
2867 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2868 leaves it alone. Component reversal etc. now happens only for IP addresses.
2871 34. The host_aton() function is supposed to be passed a string that is known
2872 to be a valid IP address. However, in the case of IPv6 addresses, it was
2873 not checking this. This is a hostage to fortune. Exim now panics and dies
2874 if the condition is not met. A case was found where this could be provoked
2875 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2876 components; fortuitously, this particular loophole had already been fixed
2877 by change 4.50/55 or 4.44/33 above.
2879 If there are any other similar loopholes, the new check in host_aton()
2880 itself should stop them being exploited. The report I received stated that
2881 data on the command line could provoke the exploit when Exim was running as
2882 exim, but did not say which command line option was involved. All I could
2883 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2884 running as the user.
2887 35. There was a buffer overflow vulnerability in the SPA authentication code
2888 (which came originally from the Samba project). I have added a test to the
2889 spa_base64_to_bits() function which I hope fixes it.
2892 36. The daemon start-up calls getloadavg() while still root for those OS that
2893 need the first call to be done as root, but it missed one case: when
2894 deliver_queue_load_max is set with deliver_drop_privilege. This is
2895 necessary for the benefit of the queue runner, because there is no re-exec
2896 when deliver_drop_privilege is set.
2898 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2899 was in use, was not putting the data itself into the right store pool;
2900 consequently, it could be overwritten for a subsequent message in the same
2901 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2904 38. Sometimes the final signoff response after QUIT could fail to get
2905 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2906 < 0 before doing a fflush(). This bug looks as though it goes back to the
2907 introduction of TLS in release 3.20, but "sometimes" must have been rare
2908 because the tests only now provoked it.
2914 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2915 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2916 Now it returns an integer. A number of calls were still expecting a T/F
2917 return. Fortuitously, in all cases, the tests worked in OK situations,
2918 which is the norm. However, things would have gone wrong on any write
2919 failures on the smtp file descriptor. This function is used when sending
2920 messages over SMTP and also when doing verify callouts.
2922 2. When Exim is called to do synchronous delivery of a locally submitted
2923 message (the -odf or -odi options), it no longer closes stderr before doing
2926 3. Implemented the mua_wrapper option.
2928 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2930 5. Implemented the functions header_remove(), header_testname(),
2931 header_add_at_position(), and receive_remove_recipient(), and exported them
2934 6. If an ACL "warn" statement specified the addition of headers, Exim already
2935 inserted X-ACL-Warn: at the start if there was no header name. However, it
2936 was not making this test for the second and subsequent header lines if
2937 there were newlines in the string. This meant that an invalid header could
2938 be inserted if Exim was badly configured.
2940 7. Allow an ACL "warn" statement to add header lines at the start or after all
2941 the Received: headers, as well as at the end.
2943 8. Added the rcpt_4xx retry error code.
2945 9. Added postmaster_mailfrom=xxx to callout verification option.
2947 10. Added mailfrom=xxxx to the callout verification option, for verify=
2950 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2951 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2953 12. Inserted some casts to stop certain compilers warning when using pointer
2954 differences as field lengths or precisions in printf-type calls (mostly
2955 affecting debugging statements).
2957 13. Added optional readline() support for -be (dynamically loaded).
2959 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2960 same clock tick as a message's arrival, so that its received time was the
2961 same as the "first fail" time on the retry record, and that message
2962 remained on the queue past the ultimate address timeout, every queue runner
2963 would try a delivery (because it was past the ultimate address timeout) but
2964 after another failure, the ultimate address timeout, which should have then
2965 bounced the address, did not kick in. This was a "< instead of <=" error;
2966 in most cases the first failure would have been in the next clock tick
2967 after the received time, and all would be well.
2969 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2970 being recognized when the domain list was tested by the match_domain
2971 condition in an expansion string.
2973 16. Added the ${str2b64: operator.
2975 17. Exim was always calling setrlimit() to set a large limit for the number of
2976 processes, without checking whether the existing limit was already
2977 adequate. (It did check for the limit on file descriptors.) Furthermore,
2978 errors from getrlimit() and setrlimit() were being ignored. Now they are
2979 logged to the main and panic logs, but Exim does carry on, to try to do its
2980 job under whatever limits there are.
2982 18. Imported PCRE 5.0.
2984 19. Trivial typo in log message " temporarily refused connection" (the leading
2987 20. If the log selector return_path_on_delivery was set and an address was
2988 redirected to /dev/null, the delivery process crashed because it assumed
2989 that a return path would always be set for a "successful" delivery. In this
2990 case, the whole delivery is bypassed as an optimization, and therefore no
2993 21. Internal re-arrangement: the function for sending a challenge and reading
2994 a response while authentication was assuming a zero-terminated challenge
2995 string. It's now changed to take a pointer and a length, to allow for
2996 binary data in such strings.
2998 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3000 23. Exim was not respecting finduser_retries when seeking the login of the
3001 uid under which it was called; it was always trying 10 times. (The default
3002 setting of finduser_retries is zero.) Also, it was sleeping after the final
3003 failure, which is pointless.
3005 24. Implemented tls_on_connect_ports.
3007 25. Implemented acl_smtp_predata.
3009 26. If the domain in control=submission is set empty, Exim assumes that the
3010 authenticated id is a complete email address when it generates From: or
3011 Sender: header lines.
3013 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3014 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3015 chown and chgrp in /bin and hostname in /usr/bin.
3017 28. Exim was keeping the "process log" file open after each use, just as it
3018 does for the main log. This opens the possibility of it remaining open for
3019 long periods when the USR1 signal hits a daemon. Occasional processlog
3020 errors were reported, that could have been caused by this. Anyway, it seems
3021 much more sensible not to leave this file open at all, so that is what now
3024 29. The long-running daemon process does not normally write to the log once it
3025 has entered its main loop, and it closes the log before doing so. This is
3026 so that log files can straightforwardly be renamed and moved. However,
3027 there are a couple of unusual error situations where the daemon does write
3028 log entries, and I had neglected to close the log afterwards.
3030 30. The text of an SMTP error response that was received during a remote
3031 delivery was being truncated at 512 bytes. This is too short for some of
3032 the long messages that one sometimes sees. I've increased the limit to
3035 31. It is now possible to make retry rules that apply only when a message has a
3036 specific sender, in particular, an empty sender.
3038 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3039 it possible to be selective about when SMTP synchronization is enforced.
3041 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3043 32. Implemented hosts_connection_nolog.
3045 33. Added an ACL for QUIT.
3047 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3050 35. Added mailbox_size and mailbox_filecount to appendfile.
3052 36. Added control = no_multiline_responses to ACLs.
3054 37. There was a bug in the logic of the code that waits for the clock to tick
3055 in the case where the clock went backwards by a substantial amount such
3056 that the microsecond fraction of "now" was more than the microsecond
3057 fraction of "then" (but the whole seconds number was less).
3059 38. Added support for the libradius Radius client library this is found on
3060 FreeBSD (previously only the radiusclient library was supported).
3066 1. When certain lookups returned multiple values in the form name=value, the
3067 quoting of the values was not always being done properly. Specifically:
3068 (a) If the value started with a double quote, but contained no whitespace,
3070 (b) If the value contained whitespace other than a space character (i.e.
3071 tabs or newlines or carriage returns) it was not quoted.
3072 This fix has been applied to the mysql and pgsql lookups by writing a
3073 separate quoting function and calling it from the lookup code. The fix
3074 should probably also be applied to nisplus, ibase and oracle lookups, but
3075 since I cannot test any of those, I have not disturbed their existing code.
3077 2. A hit in the callout cache for a specific address caused a log line with no
3078 reason for rejecting RCPT. Now it says "Previous (cached) callout
3079 verification failure".
3081 3. There was an off-by-one bug in the queryprogram router. An over-long
3082 return line was truncated at 256 instead of 255 characters, thereby
3083 overflowing its buffer with the terminating zero. As well as fixing this, I
3084 have increased the buffer size to 1024 (and made a note to document this).
3086 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3087 when Exim is waiting for an SMTP response from a remote server, Exim
3088 restarts its select() call on the socket, thereby resetting its timeout.
3089 This is not a problem when such interrupts are rare. Somebody set up a cron
3090 job to run exiwhat every 2 minutes, which is less than the normal select()
3091 timeout (5 or 10 minutes). This meant that the select() timeout never
3092 kicked in because it was always reset. I have fixed this by comparing the
3093 time when an interrupt arrives with the time at the start of the first call
3094 to select(). If more time than the timeout has elapsed, the interrupt is
3095 treated as a timeout.
3097 5. Some internal re-factoring in preparation for the addition of Sieve
3098 extensions (by MH). In particular, the "personal" test is moved to a
3099 separate function, and given an option for scanning Cc: and Bcc: (which is
3100 not set for Exim filters).
3102 6. When Exim created an email address using the login of the caller as the
3103 local part (e.g. when creating a From: or Sender: header line), it was not
3104 quoting the local part when it contained special characters such as @.
3106 7. Installed new OpenBSD configuration files.
3108 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3109 try to make them clearer.
3111 9. Callout options, other than the timeout value, were being ignored when
3112 verifying sender addresses in header lines. For example, when using
3114 verify = header_sender/callout=no_cache
3116 the cache was (incorrectly) being used.
3118 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3119 only the Cygwin environment.
3121 11. When return_path_on_delivery was set as a log selector, if different remote
3122 addresses in the same message used different return paths and parallel
3123 remote delivery occurred, the wrong values would sometimes be logged.
3124 (Whenever a remote delivery process finished, the return path value from
3125 the most recently started remote delivery process was logged.)
3127 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3128 lines when AUTH and/or TLS are in use. This is the "received protocol"
3129 field. Exim used to use "asmtp" for authenticated SMTP, without any
3130 indication (in the protocol name) for TLS use. Now it follows the RFC and
3131 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3132 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3133 names appear in log lines as well as in Received: header lines.
3135 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3136 extensions, and comparison tests, and to fix some bugs.
3138 14. Changes to the "personal" filter test:
3140 (1) The test was buggy in that it was just doing the equivalent of
3141 "contains" tests on header lines. For example, if a user's address was
3142 anne@some.where, the "personal" test would incorrectly be true for
3144 To: susanne@some.where
3146 This test is now done by extracting each address from the header in turn,
3147 and checking the entire address. Other tests that are part of "personal"
3148 are now done using regular expressions (for example, to check local parts
3149 of addresses in From: header lines).
3151 (2) The list of non-personal local parts in From: addresses has been
3152 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3153 taken from the Sieve specification recommendations.
3155 (3) If the message contains any header line starting with "List-" it is
3156 treated as non-personal.
3158 (4) The test for "circular" in the Subject: header line has been removed
3159 because it now seems ill-conceived.
3161 15. Minor typos in src/EDITME comments corrected.
3163 16. Installed latest exipick from John Jetmore.
3165 17. If headers_add on a router specified a text string that was too long for
3166 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3167 of string_sprintf() is now avoided.
3169 18. $message_body_size was not set (it was always zero) when running the DATA
3170 ACL and the local_scan() function.
3172 19. For the "mail" command in an Exim filter, no default was being set for
3173 the once_repeat time, causing a random time value to be used if "once" was
3174 specified. (If the value happened to be <= 0, no repeat happened.) The
3175 default is now 0s, meaning "never repeat". The "vacation" command was OK
3176 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3177 (I found it when inspecting the code).
3179 20. There is now an overall timeout for performing a callout verification. It
3180 defaults to 4 times the callout timeout, which applies to individual SMTP
3181 commands during the callout. The overall timeout applies when there is more
3182 than one host that can be tried. The timeout is checked before trying the
3183 next host. This prevents very long delays if there are a large number of
3184 hosts and all are timing out (e.g. when the network connections are timing
3185 out). The value of the overall timeout can be changed by specifying an
3186 additional sub-option for "callout", called "maxwait". For example:
3188 verify = sender/callout=5s,maxwait=20s
3190 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3191 to the end before writing, but this should make it even safer).
3193 22. Exim was forgetting that it had advertised PIPELINING for the second and
3194 subsequent messages on an SMTP connection. It was also not resetting its
3195 memory on STARTTLS and an internal HELO.
3197 23. When Exim logs an SMTP synchronization error within a session, it now
3198 records whether PIPELINING has been advertised or not.
3200 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3201 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3202 rather than long int.
3204 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3206 26. Added the never_mail option to autoreply.
3212 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3213 crash if the getsockname() call failed; this can happen if a connection is
3214 closed very soon after it is established. The problem was simply in the
3215 order in which certain operations were done, causing Exim to try to write
3216 to the SMTP stream before it had set up the file descriptor. The bug has
3217 been fixed by making things happen in the correct order.
3223 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3224 before the connection was closed, thus losing the rejection response.
3226 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3227 some early Solaris releases, but causes trouble in current releases where
3228 socklen_t is defined.
3230 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3233 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3236 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3237 that releasing the top of it at the end releases what was used for sub-
3238 expansions (unless the block got too big). However, discard this block if
3239 the first thing is a variable or header, so that we can use its block when
3240 it is dynamic (useful for very large $message_headers, for example).
3242 6. Lookups now cache *every* query, not just the most recent. A new, separate
3243 store pool is used for this. It can be recovered when all lookup caches are
3244 flushed. Lookups now release memory at the end of their result strings.
3245 This has involved some general refactoring of the lookup sources.
3247 7. Some code has been added to the store_xxx() functions to reduce the amount
3248 of flapping under certain conditions.
3250 8. log_incoming_interface used to affect only the <= reception log lines. Now
3251 it causes the local interface and port to be added to several more SMTP log
3252 lines, for example "SMTP connection from", and rejection lines.
3254 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3256 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3258 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3259 could be overwritten at the end of the current message (or the start of a
3260 new message if it was set in a HELO ACL). The value is now preserved for
3261 the duration of the SMTP connection.
3263 12. If a transport had a headers_rewrite setting, and a matching header line
3264 contained an unqualified address, that address was qualified, even if it
3265 did not match any rewriting rules. The underlying bug was that the values
3266 of the flags that permit the existence of unqualified sender and recipient
3267 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3268 non-local messages, and by -bnq for local messages) were not being
3269 preserved with the message after it was received.
3271 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3272 "next input=" as part of the text comprising the host identity instead of
3273 the correct text. The code was using the same buffer for two different
3274 strings. However, depending on which order the printing function evaluated
3275 its arguments, the bug did not always show up. Under Linux, for example, my
3276 test suite worked just fine.
3278 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3279 doesn't work with some older versions of Perl. It has been changed to "my",
3280 which in any case is probably the better facility to use.
3282 15. A really picky compiler found some instances of statements for creating
3283 error messages that either had too many or two few arguments for the format
3286 16. The size of the buffer for calls to the DNS resolver has been increased
3287 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3288 for addresses that have a lot of PTR records. This alleviates a problem; it
3289 does not fully solve it.
3291 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3292 buffer now truncates the list and logs the incident, which is the same
3293 action as happens when Exim is looking up a host name and its aliases.
3294 Previously in this situation something unpredictable would happen;
3295 sometimes it was "internal error: store_reset failed".
3297 18. If a server dropped the connection unexpectedly when an Exim client was
3298 using GnuTLS and trying to read a response, the client delivery process
3299 crashed while trying to generate an error log message.
3301 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3302 string, for example:
3304 warn message = H1: something\nH2: something
3306 the text was added as a single header line from Exim's point of view
3307 though it ended up OK in the delivered message. However, searching for the
3308 second and subsequent header lines using $h_h2: did not work. This has been
3309 fixed. Similarly, if a system filter added multiple headers in this way,
3310 the routers could not see them.
3312 20. Expanded the error message when iplsearch is called with an invalid key to
3313 suggest using net-iplsearch in a host list.
3315 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3316 ACLs are no longer actually imposed (and a message to that effect is
3319 22. If a "gecos" field in a passwd entry contained escaped characters, in
3320 particular, if it contained a \" sequence, Exim got it wrong when building
3321 a From: or a Sender: header from that name. A second bug also caused
3322 incorrect handling when an unquoted " was present following a character
3323 that needed quoting.
3325 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3326 was not being matched caselessly.
3328 24. Arranged for all hyphens in the exim.8 source to be escaped with
3331 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3332 properly. Recipient callout cache records were still being keyed to include
3333 the sender, even when use_sender was set false. This led to far more
3334 callouts that were necessary. The sender is no longer included in the key
3335 when use_sender is false.
3337 26. Added "control = submission" modifier to ACLs.
3339 27. Added the ${base62d: operator to decode base 62 numbers.
3341 28. dnsdb lookups can now access SRV records.
3343 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3344 the configuration file.
3346 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3347 (-v) mode. This makes the output for a verbose queue run more intelligible.
3349 31. Added a use_postmaster feature to recipient callouts.
3351 32. Added the $body_zerocount variable, containing the number of binary zero
3352 bytes in the message body.
3354 33. The time of last modification of the "new" subdirectory is now used as the
3355 "mailbox time last read" when there is a quota error for a maildir
3358 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3360 35. Added +ignore_unknown as a special item in host lists.
3362 36. Code for decoding IPv6 addresses in host lists is now included, even if
3363 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3364 address was recognized as an IP address, but was then not correctly decoded
3365 into binary, causing unexpected and incorrect effects when compared with
3372 1. Very minor rewording of debugging text in manualroute to say "list of
3373 hosts" instead of "hostlist".
3375 2. If verify=header_syntax was set, and a header line with an unqualified
3376 address (no domain) and a large number of spaces between the end of the
3377 name and the colon was received, the reception process suffered a buffer
3378 overflow, and (when I tested it) crashed. This was caused by some obsolete
3379 code that should have been removed. The fix is to remove it!
3381 3. When running in the test harness, delay a bit after writing a bounce
3382 message to get a bit more predictability in the log output.
3384 4. Added a call to search_tidyup() just before forking a reception process. In
3385 theory, someone could use a lookup in the expansion of smtp_accept_max_
3386 per_host which, without the tidyup, could leave open a database connection.
3388 5. Added the variables $recipient_data and $sender_data which get set from a
3389 lookup success in an ACL "recipients" or "senders" condition, or a router
3390 "senders" option, similar to $domain_data and $local_part_data.
3392 6. Moved the writing of debug_print from before to after the "senders" test
3395 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3396 problems for message scanning, either using a data ACL, or using
3397 local_scan() because the Received: header was not generated till after they
3398 were called (in order to set the time as the time of reception completion).
3399 I have revised the way this works. The header is now generated after the
3400 body is received, but before the ACL or local_scan() are called. After they
3401 are run, the timestamp in the header is updated.
3407 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3408 before starting a queue runner without re-exec. This happened only when
3409 deliver_drop_privilege was set or when the Exim user was set to root. The
3410 effect of the bug was that timeouts during subsequent deliveries caused
3411 crashes instead of being properly handled. The handler is now left at its
3412 default (and expected) setting.
3414 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3415 message, again when deliver_drop_privilege is set or Exim is run as root.
3416 The bug described in (1) was not present in this case, but the tidying up
3417 of the other signals was missing. I have made the two cases consistent.
3419 3. The ignore_target_hosts setting on a manualroute router was being ignored
3420 for hosts that were looked up using the /MX notation.
3422 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3425 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3426 operated on the sender address. After changing the $sender_address to <>
3427 for the sender address verify, Exim was re-instated it as the original
3428 (before rewriting) address, but remembering that it had rewritten it, so it
3429 wasn't rewriting it again. This bug also had the effect of breaking the
3430 sender address verification caching when the sender address was rewritten.
3432 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3433 This has been changed so that if the ip literal address matches
3434 ignore_target_hosts, the router declines.
3436 7. Added expansion conditions match_domain, match_address, and match_local_
3437 part (NOT match_host).
3439 8. The placeholder for the Received: header didn't have a length field set.
3441 9. Added code to Exim itself and to exim_lock to test for a specific race
3442 condition that could lead to file corruption when using MBX delivery. The
3443 issue is with the lockfile that is created in /tmp. If this file is removed
3444 after a process has opened it but before that process has acquired a lock,
3445 there is the potential for a second process to recreate the file and also
3446 acquire a lock. This could lead to two Exim processes writing to the file
3447 at the same time. The added code performs the same test as UW imapd; it
3448 checks after acquiring the lock that its file descriptor still refers to
3449 the same named file.
3451 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3452 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3455 11. Added the smtp_active_hostname option. If used, this will typically be made
3456 to depend on the incoming interface address. Because $interface_address is
3457 not set up until the daemon has forked a reception process, error responses
3458 that can happen earlier (such as "too many connections") no longer contain
3461 12. If an expansion in a condition on a "warn" statement fails because a lookup
3462 defers, the "warn" statement is abandoned, and the next ACL statement is
3463 processed. Previously this caused the whole ACL to be aborted.
3465 13. Added the iplsearch lookup type.
3467 14. Added ident_timeout as a log selector.
3469 15. Added tls_certificate_verified as a log selector.
3471 16. Added a global option tls_require_ciphers (compare the smtp transport
3472 option of the same name). This controls incoming TLS connections.
3474 17. I finally figured out how to make tls_require_ciphers do a similar thing
3475 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3476 before starting the TLS session.
3478 18. Tabs are now shown as \t in -bP output.
3480 19. If the log selector return_path_on_delivery was set, Exim crashed when
3481 bouncing a message because it had too many Received: header lines.
3483 20. If two routers both had headers_remove settings, and the first one included
3484 a superfluous trailing colon, the final name in the first list and the
3485 first name in the second list were incorrectly joined into one item (with a
3486 colon in the middle).
3492 1. Added -C and -D options to the exinext utility, mainly to make it easier
3493 to include in the automated testing, but these could be helpful when
3494 multiple configurations are in use.
3496 2. The exinext utility was not formatting the output nicely when there was
3497 an alternate port involved in the retry record key, nor when there was a
3498 message id as well (for retries that were specific to a specific message
3499 and a specific host). It was also confused by IPv6 addresses, because of
3500 the additional colons they contain. I have fixed the IPv4 problem, and
3501 patched it up to do a reasonable job for IPv6.
3503 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3504 delivery, the log line now contains "pipelined" if PIPELINING was used.
3506 4. An SMTP transport process used to panic and die if the bind() call to set
3507 an explicit outgoing interface failed. This has been changed; it is now
3508 treated in the same way as a connect() failure.
3510 5. A reference to $sender_host_name in the part of a conditional expansion
3511 that was being skipped was still causing a DNS lookup. This no longer
3514 6. The def: expansion condition was not recognizing references to header lines
3515 that used bh_ and bheader_.
3517 7. Added the _cache feature to named lists.
3519 8. The code for checking quota_filecount in the appendfile transport was
3520 allowing one more file than it should have been.
3522 9. For compatibility with Sendmail, the command line option
3530 and sets the incoming protocol and host name (for trusted callers). The
3531 host name and its colon can be omitted when only the protocol is to be set.
3532 Note the Exim already has two private options, -pd and -ps, that refer to
3533 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3534 "s", but I don't think that's a major issue.
3536 10. A number of refactoring changes to the code, none of which should affect
3539 (a) The number of logging options was getting close to filling up the
3540 32-bit word that was used as a bit map. I have split them into two classes:
3541 those that are passed in the argument to log_write(), and those that are
3542 only ever tested independently outside of that function. These are now in
3543 separate 32-bit words, so there is plenty of room for expansion again.
3544 There is no change in the user interface or the logging behaviour.
3546 (b) When building, for example, log lines, the code previously used a
3547 macro that called string_cat() twice, in order to add two strings. This is
3548 not really sufficiently general. Furthermore, there was one instance where
3549 it was actually wrong because one of the argument was used twice, and in
3550 one call a function was used. (As it happened, calling the function twice
3551 did not affect the overall behaviour.) The macro has been replaced by a
3552 function that can join an arbitrary number of extra strings onto a growing
3555 (c) The code for expansion conditions now uses a table and a binary chop
3556 instead of a serial search (which was left over from when there were very
3557 few conditions). Also, it now recognizes conditions like "pam" even when
3558 the relevant support is not compiled in: a suitably worded error message is
3559 given if an attempt is made to use such a condition.
3561 11. Added ${time_interval:xxxxx}.
3563 12. A bug was causing one of the ddress fields not to be passed back correctly
3564 from remote delivery subprocesses. The field in question was not being
3565 subsequently used, so this caused to problems in practice.
3567 13. Added new log selectors queue_time and deliver_time.
3569 14. Might have fixed a bug in maildirsizefile handling that threw up
3570 "unexpected character" debug warnings, and recalculated the data
3571 unnecessarily. In any case, I expanded the warning message to give more
3574 15. Added the message "Restricted characters in address" to the statements in
3575 the default ACL that block characters like @ and % in local parts.
3577 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3578 Three changes have been made:
3580 (a) There was a serious bug; a negative response to MAIL caused the whole
3581 recipient domain to be cached as invalid, thereby blocking all messages
3582 to all local parts at the same domain, from all senders. This bug has
3583 been fixed. The domain is no longer cached after a negative response to
3584 MAIL if the sender used is not empty.
3586 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3589 (c) A new callout option, "use_sender" has been added for people who want
3590 the modified behaviour.
3596 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3599 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3600 indeed breaks things for older releases.
3602 3. Added additional logging to the case where there is a problem reading data
3603 from a filter that is running in a subprocess using a pipe, in order to
3604 try to track down a specific problem.
3606 4. Testing facility fudge: when running in the test harness and attempting
3607 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3608 getting "No route to host". Convert this to a timeout.
3610 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3613 6. Some OS don't have socklen_t but use size_t instead. This affects the
3614 fifth argument of getsockopt() amongst other things. This is now
3615 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3616 can be set for individual OS. I have set it for SunOS5, OSF1, and
3617 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3618 some earlier ones do not.
3620 7. Change 4.30/15 was not doing the test caselessly.
3622 8. The standard form for an IPv6 address literal was being rejected by address
3623 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3624 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3625 this, as well as the form without the "IPv6" on the front (but only when
3626 address literals are enabled, of course).
3628 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3630 10. Exim crashed if a message with an empty sender address specified by -f
3631 encountered a router with an errors_to setting. This could be provoked only
3632 by a command such as
3636 where an empty string was supplied; "<>" did not hit this bug.
3638 11. Installed PCRE release 4.5.
3640 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3641 remained set. It is now erased.
3643 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3644 times from message ids (which are base 36 rather than the normal 62).
3646 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3647 were being counted as actual protocol errors, and logged if the log
3648 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3649 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3650 and DATA following a set of rejected RCPTs do not count as protocol errors.
3651 In other words, Exim assumes they were pipelined, though this may not
3652 actually be the case. Of course, in all cases the client gets an
3653 appropriate error code.
3655 15. If a lookup fails in an ACL condition, a message about the failure may
3656 be available; it is used if testing the ACL cannot continue, because most
3657 such messages specify what the cause of the deferral is. However, some
3658 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3659 that caused an old message to be retained and used if a later statement
3660 caused a defer, replacing the real cause of the deferral.
3662 16. If an IP address had so many PTR records that the DNS lookup buffer
3663 was not large enough to hold them, Exim could crash while trying to process
3664 the truncated data. It now detects and logs this case.
3666 17. Further to 4.21/58, another change has been made: if (and only if) the
3667 first line of a message (the first header line) ends with CRLF, a bare LF
3668 in a subsequent header line has a space inserted after it, so as not to
3669 terminate the header.
3671 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3672 unnecessarily, used atoi() instead of strtol(), and didn't check the
3673 termination when getting file sizes from file names by regex.
3675 19. Completely re-implemented the support for maildirsize files, in the light
3676 of a number of problems with the previous contributed implementation
3677 (4.30/29). In particular:
3679 . If the quota is zero, the maildirsize file is maintained, but no quota is
3682 . If the maildir directory does not exist, it is created before any attempt
3683 to write a maildirsize file.
3685 . The quota value in the file is just a cache; if the quota is changed in
3686 the transport, the new value overrides.
3688 . A regular expression is available for excluding directories from the
3691 20. The autoreply transport checks the characters in options that define the
3692 message's headers; it allows continued headers, but it was checking with
3693 isspace() after an embedded newline instead of explicitly looking for a
3696 21. If all the "regular" hosts to which an address was routed had passed their
3697 expiry times, and had not reached their retry times, the address was
3698 bounced, even if fallback hosts were defined. Now Exim should go on to try
3701 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3702 equivalent code in the SMTP transport. Some hosts send humungous responses
3703 to HELO/EHLO, more than 1024 it seems.
3705 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3706 gives compiler warnings in some environments. I've now done it "properly",
3709 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3710 (because of problems with the built-in one) was declared to return uschar *
3711 instead of char *, causing compiler failure.
3713 25. Fixed a file descriptor leak when processing alias/forward files.
3715 26. Fixed a minor format string issue in dbfn.c.
3717 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3719 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3720 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3723 29. When a sender address is verified, it is cached, to save repeating the test
3724 when there is more than one recipient in a message. However, when the
3725 verification involves a callout, it is possible for different callout
3726 options to be set for different recipients. It is too complicated to keep
3727 track of this in the cache, so now Exim always runs a verification when a
3728 callout is required, relying on the callout cache for the optimization.
3729 The overhead is duplication of the address routing, but this should not be
3732 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3733 to be verified with callout=postmaster, and the main callout worked but the
3734 postmaster check failed, the verification correctly failed. However, if a
3735 subsequent RCPT command asked for sender verification *without* the
3736 postmaster check, incorrect caching caused this verification also to fail,
3739 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3740 it was not caching the DNS options (qualify_single, search_parents) that
3741 were used when the lookup failed. A subsequent lookup with different
3742 options therefore always gave the same answer, though there were cases
3743 where it should not have. (Example: a "domains = !$mx_any" option on a
3744 dnslookup router: the "domains" option is always processed without any
3745 widening, but the router might have qualify_single set.) Now Exim uses the
3746 cached value only when the same options are set.
3748 32. Added John Jetmore's "exipick" utility to the distribution.
3750 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3751 than a timeout (e.g. a certificate is required, and is not provided), an
3752 Exim server now closes the connection immediately. Previously it waited for
3753 the client to close - but if the client is SSL, it seems that they each
3754 wait for each other, leading to a delay before one of them times out.
3756 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3757 maintained 0.8.x compatibility because I don't think many are using it, and
3758 it is clearly obsolete.
3760 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3763 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3764 client certificate was expired. A simple patch fixes this, though I don't
3765 understand the full logic of why the verify callback is called multiple
3768 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3769 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3770 which causes problems with some clients (such as the Certicom SSL Plus
3771 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3772 disables the coutermeasure allowing Eudora to connect."
3774 38. Exim was not checking that a write() to a log file succeeded. This could
3775 lead to Bad Things if a log got too big, in particular if it hit a file
3776 size limit. Exim now panics and dies if it cannot write to a log file, just
3777 as it does if it cannot open a log file.
3779 39. Modified OS/Makefile-Linux so that it now contains
3781 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3783 The two -D definitions ensure that Exim is compiled with large file
3784 support, which makes it possible to handle log files that are bigger than
3787 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3788 instance) a domain was checked against a named list that involved a lookup,
3789 causing $domain_data to be set, then another domain was checked against the
3790 same list, then the first domain was re-checked, the value of $domain_data
3791 after the final check could be wrong. In particular, if the second check
3792 failed, it could be set empty. This bug probably also applied to
3795 41. The strip_trailing_dot option was not being applied to the address given
3796 with the -f command-line option.
3798 42. The code for reading a message's header from the spool was incrementing
3799 $received_count, but never initializing it. This meant that the value was
3800 incorrect (doubled) while delivering a message in the same process in which
3801 it was received. In the most common configuration of Exim, this never
3802 happens - a fresh exec is done - but it can happen when
3803 deliver_drop_privilege is set.
3805 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3806 it now includes up to 150 characters of the unexpected data in the log
3809 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3810 and building data strings. The size of both of these buffers was 10 000
3811 bytes - far larger than anybody would *ever* want, thought I. Needless to
3812 say, somebody hit the limit. I have increased the maximum line length to
3813 20 000 and the maximum data length of concatenated lines to 100 000. I have
3814 also fixed two bugs, because there was no checking on these buffers. Tsk,
3815 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3816 buffer is too small.
3818 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3819 lsearch lookups. Now it does.
3821 46. When parsing a route_list item in a manualroute router, a fixed-length
3822 buffer was used for the list of hosts. I made this 1024 bytes long,
3823 thinking that nobody would ever have a list of hosts that long. Wrong.
3824 Somebody had a whole pile of complicated expansion conditions, and the
3825 string was silently truncated, leading to an expansion error. It turns out
3826 that it is easier to change to an unlimited length (owing to other changes
3827 that have happened since this code was originally written) than to build
3828 structure for giving a limitation error. The length of the item that
3829 expands into the list of hosts is now unlimited.
3831 47. The lsearch lookup could not handle data where the length of text line was
3832 more than 4095 characters. Such lines were truncated, leading to shortened
3833 data being returned. It should now handle lines of any length.
3835 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3836 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3839 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3840 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3841 no longer happens in comment lines. A list of replacements is now placed
3842 at the head of all of the source files, except those whose only change is
3843 to replace PERL_COMMAND in the very first #! line.
3845 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3846 messages on the queue, with a bottom-up merge sort, using code contributed
3847 by Michael Haardt. This should make operations like -bp somewhat faster on
3848 large queues. It won't affect queue runners, except when queue_run_in_order
3851 51. Installed eximstats 1.31 in the distribution.
3853 52. Added support for SRV lookups to the dnslookup router.
3855 53. If an ACL referred to $message_body or $message_body_end, the value was not
3856 reset for any messages that followed in the same SMTP session.
3858 54. The store-handling optimization for building very long strings was not
3859 differentiating between the different store pools. I don't think this
3860 actually made any difference in practice, but I've tidied it.
3862 55. While running the routers to verify a sender address, $sender_address
3863 was still set to the sender address. This is wrong, because when routing to
3864 send a bounce to the sender, it would be empty. Therefore, I have changed
3865 it so that, while verifying a sender address, $sender_address is set to <>.
3866 (There is no change to what happens when verifying a recipient address.)
3868 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3869 target A or AAAA records (if not already returned) without resetting the
3870 qualify_single or search_parents options of the DNS resolver. These are
3871 inappropriate in this case because the targets of MX and SRV records must
3872 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3873 target that, when qualified, matched something in the local domain. These
3874 two options are now turned off when doing these lookups.
3876 57. It seems that at least some releases of Reiserfs (which does not have the
3877 concept of a fixed number of inodes) returns zero and not -1 for the
3878 number of available inodes. This interacted badly with check_spool_inodes,
3879 which assumed that -1 was the "no such thing" setting. What I have done is
3880 to check that the total number of inodes is greater than zero before doing
3881 the test of how many are available.
3883 58. When a "warn" ACL statement has a log_message modifier, the message is
3884 remembered, and not repeated. This is to avoid a lot of repetition when a
3885 message has many recipients that cause the same warning to be written.
3886 Howewer, Exim was preserving the list of already written lines for an
3887 entire SMTP session, which doesn't seem right. The memory is now reset if a
3888 new message is started.
3890 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3891 debugging output unless log_rewrite was also set.
3893 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3894 of (int)(handle) when we know that handle contains (void *)(-1).
3896 61. The Exim daemon panic-logs an error return when it closes the incoming
3897 connection. However "connection reset by peer" seems to be common, and
3898 isn't really an error worthy of noting specially, so that particular error
3901 62. When Exim is trying to find all the local interfaces, it used to panic and
3902 die if the ioctl to get the interface flags failed. However, it seems that
3903 on at least one OS (Solaris 9) it is possible to have an interface that is
3904 included in the list of interfaces, but for which you get a failure error
3905 for this call. This happens when the interface is not "plumbed" into a
3906 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3907 failure of the "get flags" call assumes that the interface is down.
3909 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3910 makes life easier for people who are doing arithmetic on fields extracted
3911 from dates, where you often get leading zeros that should not be
3912 interpreted as octal.
3914 64. Added qualify_domain to the redirect router, to override the global
3917 65. If a pathologically long header line contained very many addresses (the
3918 report of this problem mentioned 10 000) and each of them was rewritten,
3919 Exim could use up a very large amount of memory. (It kept on making new
3920 copies of the header line as it rewrote, and never released the old ones.)
3921 At the expense of a bit more processing, the header rewriting function has
3922 been changed so that it no longer eats memory in this way.
3924 66. The generation of the Received: header has been moved from the time that a
3925 message starts to be received, to the time that it finishes. The timestamp
3926 in the Received: header should now be very close to that of the <= log
3927 line. There are two side-effects of this change:
3929 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3930 logged header lines no longer include the local Received: line, because
3931 it has not yet been created. The same applies to a copy of the message
3932 that is returned to a non-SMTP sender when a message is rejected.
3934 (b) When a filter file is tested using -bf, no additional Received: header
3935 is added to the test message. After some thought, I decided that this
3938 This change does not affect the value of $received_for. It is still set
3939 after address rewriting, but before local_scan() is called.
3941 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3943 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3944 gave an unhelpful panic error message, and a defer error. I have managed to
3945 change this behaviour so that it now rejects any supplied certificate,
3946 which seems right, as the list of acceptable certificates is empty.
3948 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3949 gave an unhelpful defer error. I have not managed to make this reject any
3950 supplied certificates, but the error message it gives is "no certificate
3951 supplied", which is not helpful.
3953 70. exigrep's output now also includes lines that are not associated with any
3954 message, but which match the given pattern. Implemented by a patch from
3955 Martin Sluka, which also tidied up the Perl a bit.
3957 71. Recipient callout verification, like sender verification, was using <> in
3958 the MAIL FROM command. This isn't really the right thing, since the actual
3959 sender may affect whether the remote host accepts the recipient or not. I
3960 have changed it to use the actual sender in the callout; this means that
3961 the cache record is now keyed on a recipient/sender pair, not just the
3962 recipient address. There doesn't seem to be a real danger of callout loops,
3963 since a callout by the remote host to check the sender would use <>.
3964 [SEE ABOVE: changed after hitting problems.]
3966 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3967 temporary errors. However, in the case of such a code being given after
3968 the end of a data transmission (i.e. after ".") Exim was failing to write
3969 a retry record for the message. (Yes, there was some broken host that was
3970 actually sending 8xx at this point.)
3972 73. An unknown lookup type in a host list could cause Exim to panic-die when
3973 the list was checked. (An example that provoked this was putting <; in the
3974 middle of a list instead of at the start.) If this happened during a DATA
3975 ACL check, a -D file could be left lying around. This kind of configuration
3976 error no longer causes Exim to die; instead it causes a defer errror. The
3977 incident is still logged to the main and panic logs.
3979 74. Buglet left over from Exim 3 conversion. The message "too many messages
3980 in one connection" was written to the rejectlog but not the mainlog, except
3981 when address rewriting (yes!) was being logged.
3983 75. Added write_rejectlog option.
3985 76. When a system filter was run not as root (that is, when system_filter_user
3986 was set), the values of the $n variables were not being returned to the
3987 main process; thus, they were not subsequently available in the $sn
3990 77. Added +return_path_on_delivery log selector.
3992 78. A connection timeout was being treated differently from recipients deferred
3993 when testing hosts_max_try with a message that was older than the host's
3994 retry timeout. (The host should not be counted, thus allowing all hosts to
3995 be tried at least once before bouncing.) This may have been the cause of an
3996 occasionally reported bug whereby a message would remain on the queue
3997 longer than the retry timeout, but would be bounced if a delivery was
3998 forced. I say "may" because I never totally pinned down the problem;
3999 setting up timeout/retry tests is difficult. See also the next item.
4001 79. The ultimate address timeout was not being applied to errors that involved
4002 a combination of host plus message (for example, a timeout on a MAIL
4003 command). When an address resolved to a number of possible hosts, and they
4004 were not all tried for each delivery (e.g. because of hosts_max_try), a
4005 message could remain on the queue longer than the retry timeout.
4007 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4010 81. Fixed an obscure SMTP outgoing bug which required at least the following
4011 conditions: (a) there was another message waiting for the same server;
4012 (b) the server returned 5xx to all RCPT commands in the first message so
4013 that the message was not completed; (c) the server dropped the connection
4014 or gave a negative response to the RSET that Exim sends to abort the
4015 transaction. The observed case was a dropped connection after DATA that had
4016 been sent in pipelining mode. That is, the server had advertised PIPELINING
4017 but was not implementing it correctly. The effect of the bug was incorrect
4018 behaviour, such as trying another host, and this could lead to a crash.
4024 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4025 and daemon.c were passed as pointers to ints; they should have been
4026 pointers to socklen_t variables (which are typically unsigned ints).
4028 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4031 3. Fixed a really odd bug that affected only the testing scheme; patching a
4032 certain fixed string in the binary changed the value of another string that
4033 happened to be identical to the end of the original first string.
4035 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4036 name", it returns that address as the IP address. On some operating
4037 systems (e.g. Solaris), it also passes back the IP address string as the
4038 "host name". However, on others (e.g. Linux), it passes back an empty
4039 string. Exim wasn't checking for this, and was changing the host name to an
4040 empty string, assuming it had been canonicized.
4042 5. Although rare, it is permitted to have more than one PTR record for a given
4043 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4044 all the names associated with an address, because they do in Solaris.
4045 However, it seems that they do not in Linux for data that comes from the
4046 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4047 I found this out when I moved to a new Linux workstation and tried to run
4048 the Exim test suite.
4050 To get round this problem I have changed the code so that it now does its
4051 own call to the DNS to look up PTR records when searching for a host name.
4052 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4053 addresses that are only in /etc/hosts are still found.
4055 This behaviour is, however, controlled by an option called host_lookup_
4056 order, which defaults to "bydns:byaddr". If people want to use the other
4057 order, or indeed, just use one or the other means of lookup, they can
4058 specify it in this variable.
4060 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4061 some operating systems, this comes back from gethostbyaddr() as an empty
4062 string, and this is what Exim used to test for. However, it seems that in
4063 other systems, "." is yielded. Exim now tests for this case too.
4065 7. The values of check_spool_space and check_log_space are now held internally
4066 as a number of kilobytes instead of an absolute number of bytes. If a
4067 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4068 kilobyte. This means that much larger values can be stored.
4070 8. Exim monitor: an attempt to get the action menu when not actually pointing
4071 at a message produces an empty menu entitled "No message selected". This
4072 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4073 no entries in it ("Shell widget menu has zero width and/or height"). So I
4074 have added a single, blank menu entry in this case.
4076 9. Added ${quote_local_part.
4078 10. MIME decoding is now applied to the contents of Subject: header lines when
4081 11. Now that a reference to $sender_host_address automatically causes a reverse
4082 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4083 host lookup before query-style lookups in lists that might use this
4084 variable. This has therefore been abolished, and the "net-" prefix is no
4085 longer necessary for query-style lookups.
4087 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4088 have been a typo for LFLAGS, so it has been changed.
4090 13. The install script calls Exim with "-C /dev/null" in order to find the
4091 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4092 to be output. Howeve, since Exim outputs its version number before the
4093 error, it didn't break the script. It just looked ugly. I fixed this by
4094 always allowing "-C /dev/null" if the caller is root.
4096 14. Ignore overlarge ACL variable number when reading spool file - insurance
4097 against a later release with more variables having written the file.
4099 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4100 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4101 this, as well as the form without the "IPv6" on the front.
4103 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4104 OS/Makefile-Darwin file.
4106 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4107 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4109 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4110 variable. However, some OS use ints for inodes. I've added cast to long int
4111 to get rid of the compiler warning.
4113 19. I had forgotten to lock out "/../" in configuration file names when
4114 ALT_CONFIG_PREFIX was set.
4116 20. Routers used for verification do not need to specify transports. However,
4117 if such a router generated a host list, and callout was configured, Exim
4118 crashed, because it could not find a port number from the (non-existent)
4119 transport. It now assumes port 25 in this circumstance.
4121 21. Added the -t option to exigrep.
4123 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4124 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4125 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4126 them is set, LOOKUP_LSEARCH is forced.
4128 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4130 24. Added sender and host information to the "rejected by local_scan()" log
4131 line; previously there was no indication of these.
4133 25. Added .include_if_exists.
4135 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4136 Linux. It turns out that not all file systems support this. Apparently some
4137 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4138 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4139 not supported on the file descriptor, is now ignored when Exim is trying to
4140 sync a directory. This applies only to Linux.
4142 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4144 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4145 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4148 29. Added support for maildirsize files from supplied patch (modified a bit).
4150 30. The use of :fail: followed by an empty string could lead Exim to respond to
4151 sender verification failures with (e.g.):
4153 550 Verification failed for <xxx>
4154 550 Sender verify failed
4156 where the first response line was missing the '-' that indicates it is not
4157 the final line of the response.
4159 31. The loop for finding the name of the user that called Exim had a hardwired
4160 limit of 10; it now uses the value of finduser_retries, which is used for
4161 all other user lookups.
4163 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4166 33. Exim was neglecting to zero errno before one call of strtol() when
4167 expanding a string and expecting an integer value. On some systems this
4168 resulted in spurious "integer overflow" errors. Also, it was casting the
4169 result into an int without checking.
4171 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4172 did not work. The code looks as if it has *never* worked, though it appears
4173 to have been documented since at least releast 1.62. I have made it work.
4175 35. The "timeout_DNS" error in retry rules, also documented since at least
4176 1.62, also never worked. As it isn't clear exactly what this means, and
4177 clearly it isn't a major issue, I have abolished the feature by treating it
4178 as "timeout", and writing a warning to the main and panic logs.
4180 36. The display of retry rules for -brt wasn't always showing the error code
4183 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4184 timeout_connect_A, timeout_connect_MX.
4186 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4187 to the empty sender.
4189 39. The daemon was not analysing the content of -oX till after it had closed
4190 stderr and disconnected from the controlling terminal. This meant that any
4191 syntax errors were only noted on the panic log, and the return code from
4192 the command was 0. By re-arranging the code a little, I've made the
4193 decoding happen first, so such errors now appear on stderr, and the return
4194 code is 1. However, the actual setting up of the sockets still happens in
4195 the disconnected process, so errors there are still only recorded on the
4198 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4199 connections (as happens on some IP stacks) was logged at start up time as
4200 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4201 it from "IPv6 and IPv4", which means that two separate sockets are being
4204 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4205 says whether AF_INET or AF_INET6 was passed as an argument.
4207 42. Exiwhat output was messed up when time zones were included in log
4210 43. Exiwhat now gives more information about the daemon's listening ports,
4211 and whether -tls-on-connect was used.
4213 44. The "port" option of the smtp transport is now expanded.
4215 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4216 silently ignored. Now an error message is written to the main and panic
4219 46. There's a new ACL modifier called "logwrite" which writes to a log file
4220 as soon as it is encountered.
4222 47. Added $local_user_uid and $local_user_gid at routing time.
4224 48. Exim crashed when trying to verify a sender address that was being
4227 49. Exim was recognizing only a space character after ".include". It now also
4228 recognizes a tab character.
4230 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4231 extracting the relevant information from the specification. The man page no
4232 longer contains scrambled data for the -d option, and I've added a section
4233 at the front about calling Exim under different names.
4235 51. Added "extra_headers" argument to the "mail" command in filter files.
4237 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4240 53. Installed eximstats 1.29.
4242 54. Added transport_filter_timeout as a generic transport option.
4244 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4245 Cc: header lines. This was required by RFC 822, but it not required by RFC
4248 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4249 incoming messages that did not have them. Now it does so only if the
4250 message originates locally, that is, if there is no associated remote host
4251 address. When Resent- header lines are present, this applies to the Resent-
4252 lines rather than the non-Resent- lines.
4254 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4255 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4257 58. Messages for configuration errors now include the name of the main
4258 configuration file - useful now that there may be more than one file in a
4259 list (.included file names were always shown).
4261 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4262 for those rare installations that do not start the daemon as root or run it
4263 setuid root. I've cut out the call to initgroups() if the daemon is not
4266 60. The Exim user and group can now be bound into the binary as text strings
4267 that are looked up at the start of Exim's processing.
4269 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4271 62. Added $mailstore_basename variable.
4273 63. Installed patch to sieve.c from Michael Haardt.
4275 64. When Exim failed to open the panic log after failing to open the main log,
4276 the original message it was trying to log was written to stderr and debug
4277 output, but if they were not available (the usual case in production), it
4278 was lost. Now it is written to syslog before the two lines that record the
4279 failures to open the logs.
4281 65. Users' Exim filters run in subprocesses under the user's uid. It is
4282 possible for a "deliver" command or an alias in a "personal" command to
4283 provoke an address rewrite. If logging of address rewriting is configured,
4284 this fails because the process is not running as root or exim. There may be
4285 a better way of dealing with this, but for the moment (because 4.30 needs
4286 to be released), I have disabled address rewrite logging when running a
4287 filter in a non-root, non-exim process.
4293 1. The buildconfig auxiliary program wasn't quoting the value set for
4294 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4295 not defined. This bug was masked in 4.22 by the effect that was fixed in
4298 2. Some messages that were rejected after a message id was allocated were
4299 shown as "incomplete" by exigrep. It no longer does this for messages that
4300 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4302 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4303 have allow_domain_literals set, the ID did not get logged in the <= line.
4304 Domain literals are now always recognized in Message-ID: header lines.
4306 4. The first argument for a ${extract expansion item is the key name or field
4307 number. Leading and trailing spaces in this item were not being ignored,
4308 causing some misleading effects.
4310 5. When deliver_drop_privilege was set, single queue runner processes started
4311 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4312 same command in the process it spins off) were not dropping privilege.
4314 6. When the daemon running as "exim" started a queue runner, it always
4315 re-executed Exim in the spun-off process. This is a waste of effort when
4316 deliver_drop_privilege is set. The new process now just calls the
4317 queue-runner function directly.
4323 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4326 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4327 ignored. Though the use of -oP was forcing the writing of a pid file, it
4328 was always written to the default place.
4330 3. If the message "no IP address found for host xxxx" is generated during
4331 incoming verification, it is now followed by identification of the incoming
4332 connection (so you can more easily find what provoked it).
4334 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4336 5. Added some features to "harden" Exim a bit more against certain attacks:
4338 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4339 be put in Local/Makefile. This is like the never_users runtime option,
4340 but it cannot be overridden. The default setting is "root".
4342 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4343 prefix string with which any file named in a -C command line option
4346 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4347 is retained for -C and -D only if the caller of Exim is root. Without
4348 it, the exim user may also use -C and -D and retain privilege.
4350 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4351 command line option is disabled.
4353 6. Macro names set by the -D option must start with an upper case letter, just
4354 like macro names defined in the configuration file.
4356 7. Added "dereference=" facility to LDAP.
4358 8. Two instances of the typo "uknown" in the source files are fixed.
4360 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4361 the Configure-Makefile script screwed up while processing it.
4363 10. Incorporated PCRE 4.4.
4365 11. The SMTP synchronization check was not operating right at the start of an
4366 SMTP session. For example, it could not catch a HELO sent before the client
4367 waited for the greeting. There is now a check for outstanding input at the
4368 point when the greeting is written. Because of the duplex, asynchronous
4369 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4370 way, but not yet received, when the check is performed.
4372 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4373 on TCP/IP sockets, because this apparently causes some broken clients to
4376 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4377 unchanged) from the Cygwin maintainer.
4379 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4380 was supported instead of testing for mbx. Effectively a typo.
4382 15. The spa authenticator server code was not checking that the input it
4383 received was valid base64.
4385 16. The debug output line for the "set" modifier in ACLs was not showing the
4386 name of the variable that was being set.
4388 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4390 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4391 lookup on its own. Something else was needed to trigger the lookup. For
4392 example, a match in host_lookup or the need for a host name in a host list.
4393 Now, if $sender_host_name is referenced and the host name has not yet been
4394 looked up, a lookup is performed. If the lookup fails, the variable remains
4395 empty, and $host_lookup_failed is set to "1".
4397 19. Added "eqi" as a case-independent comparison operator.
4399 20. The saslauthd authentication condition could segfault if neither service
4400 nor realm was specified.
4402 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4403 error message that was logged was misleading, and incoming SMTP
4404 connections were dropped. The message is now more accurate, and temporary
4405 errors are given to SMTP connections.
4407 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4408 (except RSET) with a 421 error, until QUIT is received. However, it was
4409 failing to send a response to QUIT.
4411 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4412 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4413 ACL is now run after the helo_try_verify_hosts code.
4415 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4416 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4417 case-independent, so other case variants are also recognized). Apparently
4418 some systems use these upper case variants.
4420 25. If more than two messages were waiting for the same host, and a transport
4421 filter was specified for the transport, Exim sent two messages over the
4422 same TCP/IP connection, and then failed with "socket operation on non-
4423 socket" when it tried to send the third.
4425 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4427 27. The extern definition of crypt16() in expand.c was not being excluded when
4428 the OS had its own crypt16() function.
4430 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4431 as a preferred synonym for return_size_limit, both as an option and as an
4434 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4436 30. Changed the default configuration ACL to relax the local part checking rule
4437 for addresses that are not in any local domains. For these addresses,
4438 slashes and pipe symbols are allowed within local parts, but the sequence
4439 /../ is explicitly forbidden.
4441 31. SPA server authentication was not clearing the challenge buffer before
4444 32. log_message in a "warn" ACL statement was writing to the reject log as
4445 well as to the main log, which contradicts the documentation and doesn't
4446 seem right (because no rejection is happening). So I have stopped it.
4448 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4449 However, I am unable to do any testing of this.
4451 34. Fixed an infelicity in the appendfile transport. When checking directories
4452 for a mailbox, to see if any needed to be created, it was accidentally
4453 using path names with one or more superfluous leading slashes; tracing
4454 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4456 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4457 contents are added to the log line that is written for every discarded
4458 recipient. (Previously a log_message setting was ignored.)
4460 36. The ${quote: operator now quotes the string if it is empty.
4462 37. The install script runs exim in order to find its version number. If for
4463 some reason other than non-existence or emptiness, which it checks, it
4464 could not run './exim', it was installing it with an empty version number,
4465 i.e. as "exim-". This error state is now caught, and the installation is
4468 38. An argument was missing from the function that creates an error message
4469 when Exim fails to connect to the socket for saslauthd authentication.
4470 This could cause Exim to crash, or give a corrupted message.
4472 39. Added isip, isip4, and isip6 to ${if conditions.
4474 40. The ACL variables $acl_xx are now saved with the message, and can be
4475 accessed later in routers, transports, and filters.
4477 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4478 strings in the file are not string-expanded.
4480 42. If a MAIL command specified a SIZE value that was too large to fit into an
4481 int variable, the check against message_size_limit failed. Such values are
4482 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4483 day this will have to be increased, but I don't think I want to be around
4484 when emails are that large.
4491 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4492 iconv() is not standard in FreeBSD.
4494 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4495 IPv6 enabled. The observed symptom was a segmentation fault on return from
4496 the function os_common_find_running_interfaces() in src/os.c.
4498 3. In the check_special_case() function in daemon.c I had used "errno" as an
4499 argument name, which causes warnings on some systems. This was basically a
4500 typo, since it was named "eno" in the comments!
4502 4. The code that waits for the clock to tick (at a resolution of some fraction
4503 of a second) so as to ensure message-id uniqueness was always waiting for
4504 at least one whole tick, when it could have waited for less. [This is
4505 almost certainly not relevant at current processor speeds, where it is
4506 unlikely to ever wait at all. But we try to future-proof.]
4508 5. The function that sleeps for a time interval that includes fractions of a
4509 second contained a race. It did not block SIGALRM between setting the
4510 timer, and suspending (a couple of lines later). If the interval was short
4511 and the sigsuspend() was delayed until after it had expired, the suspension
4512 never ended. On busy systems this could lead to processes getting stuck for
4515 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4516 process, before it forks any delivery processes. The open lookup caching
4517 mechanism meant that the open file or database connection was passed into
4518 the delivery process. The problem was that delivery processes always tidy
4519 up cached lookup data. This could cause a problem for the next delivery
4520 process started by the queue runner, because the external queue runner
4521 process does not know about the closure. So the next delivery process
4522 still has data in the lookup cache. In the case of a file lookup, there was
4523 no problem because closing a file descriptor in a subprocess doesn't affect
4524 the parent. However, if the lookup was caching a connection to a database,
4525 the connection was closed, and the second delivery process was likely to
4526 see errors such as "PGSQL: query failed: server closed the connection
4527 unexpectedly". The problem has been fixed by closing all cached lookups
4528 in a queue runner before running a delivery process.
4530 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4531 seem to have the "const" qualifier which it has on other OS. I've
4534 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4535 *of the same type* (client or server) with the same public name that an
4536 error should be diagnosed.
4538 9. When Exim looked up a host name for an IP address, but failed to find the
4539 original IP address when looking up the host name (a safety check), it
4540 output the message "<ip address> does not match any IP for NULL", which was
4541 confusing, to say the least. The bug was that the host name should have
4542 appeared instead of "NULL".
4544 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4545 user that is built into the binary, and the -C or -D options is used, root
4546 privilege is dropped before the configuration file is read. In addition,
4547 logging is switched to stderr instead of the normal log files. If the
4548 configuration then re-defines the Exim user, the unprivileged environment
4549 is probably not what is expected, so Exim logs a panic warning message (but
4552 However, if deliver_drop_privilege is set, the unprivileged state may well
4553 be exactly what is intended, so the warning has been cut out in that case,
4554 and Exim is allowed to try to write to its normal log files.
4560 1. smtp_return_error_details was not giving details for temporary sender
4561 or receiver verification errors.
4563 2. Diagnose a configuration error if two authenticators have the same public
4566 3. Exim used not to create the message log file for a message until the first
4567 delivery attempt. This could be confusing when incoming messages were held
4568 for policy or load reasons. The message log file is now created at the time
4569 the message is received, and an initial "Received" line is written to it.
4571 4. The automatically generated man page for command line options had a minor
4572 bug that caused no ill effects; however, a more serious problem was that
4573 the procedure for building the man page automatically didn't always
4574 operate. Consequently, release 4.20 contains an out-of-date version. This
4575 shouldn't happen again.
4577 5. When building Exim with embedded Perl support, the script that builds the
4578 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4579 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4581 6. The freeze_tell option was not being used for messages that were frozen on
4582 arrival, either by an ACL or by local_scan().
4584 7. Added the smtp_incomplete_transaction log selector.
4586 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4587 was accepting AUTH without a new EHLO.
4589 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4590 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4593 10. Logging of TCP/IP connections (when configured) now happens in the main
4594 daemon process instead of the child process, so that the TCP/IP connection
4595 count is more accurate (but it can never be perfect).
4597 11. The use of "drop" in a nested ACL was not being handled correctly in the
4598 outer ACL. Now, if condition failure induced by the nested "drop" causes
4599 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4600 or "require"), the connection is dropped.
4602 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4603 that yield "discard" can now be used with an "accept" or a "discard" verb,
4604 but an error is generated for any others (because I can't see a useful way
4605 to define what should happen).
4607 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4608 are now processed in the same way as lines in the Exim configuration file.
4609 In particular, continuation lines are supported.
4611 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4613 15. Added -ti meaning -t -i.
4615 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4616 domains, and warn by logging if others are found.
4618 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4619 returned from GIFCONF when Exim is trying to find the list of interfaces on
4620 a host. The code in os.c has been modified to copy each ifreq to an aligned
4621 structure in all cases.
4623 Also, in some cases, the returned ifreq's were being copied to a 'struct
4624 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4625 means the last couple of bytes of an IPv6 address could be chopped if the
4626 ifreq contained only a normal sockaddr (14 bytes storage).
4628 18. Named domain lists were not supported in the hosts_treat_as_local option.
4629 An entry such as +xxxx was not recognized, and was treated as a literal
4632 19. Ensure that header lines added by a DATA ACL are included in the reject log
4633 if the ACL subsequently rejects the message.
4635 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4636 MD5 (which is deprecated).
4638 21. When testing a filter file using -bf, Exim was writing a message when it
4639 took the sender from a "From " line in the message, but it was not doing so
4640 when it took $return_path from a Return-Path: header line. It now does.
4642 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4643 with a valid header line field name (a series of printing characters
4644 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4646 23. Changed "disc" in the source to "disk" to conform to the documentation and
4647 the book and for uniformity.
4649 24. Ignore Sendmail's -Ooption=value command line item.
4651 25. When execve() failed while trying to run a command in a pipe transport,
4652 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4653 could be confused with a return value of 69 from the command itself. This
4654 has been changed to 127, the value the shell returns if it is asked to run
4655 a non-existent command. The wording for the related log line suggests a
4656 non-existent command as the problem.
4658 26. If received_header_text expands to an empty string, do not add a Received:
4659 header line to the message. (Well, it adds a token one on the spool, but
4660 marks it "old" so that it doesn't get used or transmitted.)
4662 27. Installed eximstats 1.28 (addition of -nt option).
4664 28. There was no check for failure on the call to getsockname() in the daemon
4665 code. This can fail if there is a shortage of resources on the system, with
4666 ENOMEM, for example. A temporary error is now given on failure.
4668 29. Contrary to the C standard, it seems that in some environments, the
4669 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4670 program. Exim now does this explicitly; it affects the formatting of
4671 timestamps using strftime().
4673 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4674 complaints. I've now initialized all the variables, to avoid this.
4676 32. Header lines added by a system filter were not being "seen" during
4677 transport-time rewrites.
4679 33. The info_callback() function passed to OpenSSL is set up with type void
4680 (*)(SSL *, int, int), as described somewhere. However, when calling the
4681 function (actually a macro) that sets it up, the type void(*)() is
4682 expected. I've put in a cast to prevent warnings from picky compilers.
4684 34. If a DNS black list lookup found a CNAME record, but there were no A
4685 records associated with the domain it pointed at, Exim crashed.
4687 35. If a DNS black list lookup returned more than one A record, Exim ignored
4688 all but the first. It now scans all returned addresses if a particular IP
4689 value is being sought. In this situation, the contents of the
4690 $dnslist_value variable are a list of all the addresses, separated by a
4693 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4694 to accept a host name and all its aliases if the forward lookup for any of
4695 them yielded the IP address of the incoming connection. Now it accepts only
4696 those names whose forward lookup yields the correct IP address. Any other
4697 names are discarded. This closes a loophole whereby a rogue DNS
4698 administrator could create reverse DNS records to break through a
4699 wildcarded host restriction in an ACL.
4701 37. If a user filter or a system filter that ran in a subprocess used any of
4702 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4703 the wrong values were passed to the pipe command ($thisaddress had the
4704 value of $0, $0 had the value of $1, etc). This bug was introduced by
4705 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4707 38. Improved the line breaking for long SMTP error messages from ACLs.
4708 Previously, if there was no break point between 40 and 75 characters, Exim
4709 left the rest of the message alone. Two changes have been made: (a) I've
4710 reduced the minimum length to 35 characters; (b) if it can't find a break
4711 point between 35 and 75 characters, it looks ahead and uses the first one
4712 that it finds. This may give the occasional overlong line, but at least the
4713 remaining text gets split now.
4715 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4716 file descriptors might be low, and that setting 1000 would always raise it.
4717 It turns out that in some environments, the limit is already over 1000 and
4718 that lowering it causes trouble. So now Exim takes care not to decrease it.
4720 40. When delivering a message, the value of $return_path is set to $sender_
4721 address at the start of routing (routers may change the value). By an
4722 oversight, this default was not being set up when an address was tested by
4723 -bt or -bv, which affected the outcome if any router or filter referred to
4726 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4727 log only when "message" or "log_message" are set. However, if one of the
4728 conditions was an address verification, or a call to a nested ACL, the
4729 messages generated by the underlying test were being passed through. This
4730 no longer happens. The underlying message is available in $acl_verify_
4731 message for both "message" and "log_message" expansions, so it can be
4732 passed through if needed.
4734 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4735 new expansion $bh_ to give the encoded byte string without charset
4736 translation. Translation happens only if iconv() is available; HAVE_ICONV
4737 indicates this at build time. HEADERS_CHARSET gives the charset to
4738 translate to; headers_charset can change it in the configuration, and
4739 "headers charset" can change it in an individual filter file.
4741 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4742 that creates RFC 2047 encoded "words" labels them as that charset instead
4743 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4744 expansion operator; (ii) when Exim creates a From: line for a local
4745 message; (iii) when a header line is rewritten to include a "phrase" part.
4747 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4748 buggy, causing it to skip the first lines of messages whose message ID
4749 ended in 'D'. This would not have bitten before Exim release 4.14, because
4750 message IDs were unlikely to end in 'D' before then. The effect was to have
4751 incorrect size information for certain domains.
4753 45. #include "config.h" was missing at the start of the crypt16.c module. This
4754 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4757 46. If there was a timeout during a "random" callout check, Exim treated it as
4758 a failure of the random address, and carried on sending RSET and the real
4759 address. If the delay was just some slowness somewhere, the response to the
4760 original RCPT would be taken as a response to RSET and so on, causing
4761 mayhem of various kinds.
4763 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4764 when I implemented it. It didn't allow for the fact that some option values
4765 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4766 the right test for positive values.
4768 48. Domain names in DNS records are case-independent. Exim always looks them up
4769 in lower case. Some resolvers return domain names in exactly the case they
4770 appear in the zone file, that is, they may contain uppercase letters. Not
4771 all resolvers do this - some return always lower case. Exim was treating a
4772 change of case by a resolver as a change of domain, similar to a widening
4773 of a domain abbreviation. This triggered its re-routing code and so it was
4774 trying to route what was effectively the same domain again. This normally
4775 caused routing to fail (because the router wouldn't handle the domain
4776 twice). Now Exim checks for this case specially, and just changes the
4777 casing of the domain that it ultimately uses when it transmits the message
4780 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4783 50. If a filter generated a file delivery with a non-absolute name (possible if
4784 no home directory exists for the router), the forbid_file option was not
4787 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4788 the existing equality matching.
4790 52. Exim was using ints instead of ino_t variables in some places where it was
4791 dealing with inode numbers.
4793 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4794 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4795 called TMPDIR, and if it finds it is different, it changes its value.
4797 54. The smtp_printf() function is now made available to local_scan() so
4798 additional output lines can be written before returning. There is also an
4799 smtp_fflush() function to enable the detection of a dropped connection.
4800 The variables smtp_input and smtp_batched_input are exported to
4803 55. Changed the default runtime configuration: the message "Unknown user"
4804 has been removed from the ACL, and instead placed on the localuser router,
4805 using the cannot_route_message feature. This means that any verification
4806 failures that generate their own messages won't get overridden. Similarly,
4807 the "Unrouteable address" message that was in the ACL for unverifiable
4808 relay addresses has also been removed.
4810 56. Added hosts_avoid_esmtp to the smtp transport.
4812 57. The exicyclog script was not checking for the esoteric option
4813 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4814 will work only if exicyclog is run under the appropriate euid.
4816 58. Following a discussion on the list, the rules by which Exim recognises line
4817 endings on incoming messages have been changed. The -dropcr and drop_cr
4818 options are now no-ops, retained only for backwards compatibility. The
4819 following line terminators are recognized: LF CRLF CR. However, special
4820 processing applies to CR:
4822 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4823 nor a local message in the state where . is a terminator.
4825 (ii) If a bare CR is encountered in a header line, an extra space is added
4826 after the line terminator so as not to end the header. The reasoning
4827 behind this is that bare CRs in header lines are most likely either
4828 to be mistakes, or people trying to play silly games.
4830 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4831 was being incorrectly given as 18 bytes larger than it should have been.
4832 This is a VOB (very old bug).
4834 60. This may never have affected anything current, but just in case it has:
4835 When the local host is found other than at the start of a list of hosts,
4836 the local host, those with the same MX, and any that follow, are discarded.
4837 When the list in question was part of a longer list of hosts, the following
4838 hosts (not currently being processed) were also being discarded. This no
4839 longer happens. I'm not sure if this situation could ever has previously
4842 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4845 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4846 header. This is something that is recommended in a new Internet Draft, and
4847 is something that is documented as being done by Sendmail. There are two
4848 possible values. For messages generated by the autoreply transport, Exim
4851 Auto-Submitted: auto-replied
4853 whereas for all other generated messages (e.g. bounces) it adds
4855 Auto-Submitted: auto-generated
4857 63. The "personal" condition in filters now includes a test for the
4858 Auto-Submitted: header. If it contains the string "auto-" the message it
4859 not considered personal.
4861 64. Added rcpt_include_affixes as a generic transport option.
4863 65. Added queue_only_override (default true).
4865 66. Added the syslog_duplication option.
4867 67. If what should have been the first header line of a message consisted of
4868 a space followed by a colon, Exim was mis-interpreting it as a header line.
4869 It isn't of course - it is syntactically invalid and should therefore be
4870 treated as the start of the message body. The misbehaviour could have
4871 caused a number of strange effects, including loss of data in subsequent
4872 header lines, and spool format errors.
4874 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4875 client host had authenticated. This control can now be exercised by an ACL
4876 for more flexibility.
4878 69. By default, callouts do not happen when testing with -bh. There is now a
4879 variant, -bhc, which does actually run the callout code, including
4880 consulting and updating the callout cache.
4882 70. Added support for saslauthd authentication, courtesy of Alexander
4885 71. If statvfs() failed on the spool or log directories while checking their
4886 size for availability, Exim confusingly gave the error "space shortage".
4887 Furthermore, in debugging mode it crashed with a floating point exception.
4888 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4889 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4890 serious problem, Exim now writes to the main and panic logs when this
4891 happens, with details of the failure. It then refuses to accept the
4892 incoming message, giving the message "spool directory problem" or "log
4893 directory problem" with a 421 code for SMTP messages.
4895 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4896 0, 1, and 2 exist, because some OS complain for execs without them (see
4897 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4898 descriptors. However, the code omitted to check that the open succeeded,
4899 causing mysterious errors if for some reason the permissions on /dev/null
4900 got screwed. Now Exim writes a message to the main and panic logs, and
4901 bombs out if it can't open /dev/null.
4903 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4904 interact so that it is all more flexible. It is supposed to remain
4905 backwards compatible. Also added extra_local_interfaces.
4907 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4908 to bomb out with an assertion failure - to the client this appears as a
4909 connection drop. This problem occurs in the part of the code that was taken
4910 from the Samba project. Fortunately, the assertion is in a very simple
4911 function, so I have fixed this by reproducing the function inline in the
4912 one place where it is called, and arranging for authentication to fail
4913 instead of killing the process with assert().
4915 75. The SPA client code was not working when the server requested OEM rather
4916 than Unicode encoding.
4918 76. Added code to make require_files with a specific uid setting more usable in
4919 the case where statting the file as root fails - usually a non-root-mounted
4920 NFS file system. When this happens and the failure is EACCES, Exim now
4921 forks a subprocess and does the per-uid checking as the relevant uid.
4923 77. Added process_log_path.
4925 78. If log_file_path was not explicitly set, a setting of check_log_space or
4926 check_log_inodes was ignored.
4928 79. If a space check for the spool or log partitions fails, the incident is now
4929 logged. Of course, in the latter case the data may get lost...
4931 80. Added the %p formatting code to string_format() so that it can be used to
4932 print addresses in debug_print(). Adjusted all the address printing in the
4933 debugging in store.c to use %p rather than %d.
4935 81. There was a concern that a line of code in smtp_in.c could overflow a
4936 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4937 initially expressed, the concern was not well-founded, because trailing
4938 spaces are removed early. However, if the trailing spaces were followed by
4939 a NULL, they did not get removed, so the overflow was possible. Two fixes
4942 (a) I re-wrote the offending code in a cleaner fashion.
4943 (b) If an incoming SMTP command contains a NULL character, it is rejected
4946 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4947 runs initgroups(), so that if the Exim user is in any additional groups,
4948 they will be used during message reception.
4954 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob