1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.548 2008/05/16 12:22:08 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: Improvedment to Dovecot authenticator handling.
55 Patch provided by Jan Srzednicki
63 TK/01 Add preliminary DKIM support. Currently requires a forked version of
64 ALT-N's libdkim that I have put here:
65 http://duncanthrax.net/exim-experimental/
67 Note to Michael Haardt: I had to rename some vars in sieve.c. They
68 were called 'true' and it seems that C99 defines that as a reserved
69 keyword to be used with 'bool' variable types. That means you could
70 not include C99-style headers which use bools without triggering
71 build errors in sieve.c.
73 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
74 as mailq or other aliases. Changed the --help handling significantly
75 to do whats expected. exim_usage() emits usage/help information.
77 SC/01 Added the -bylocaldomain option to eximstats.
79 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
81 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
83 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
89 PH/01 Another patch from the Sieve maintainer.
91 PH/02 When an IPv6 address is converted to a string for single-key lookup
92 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
93 dots are used instead of colons so that keys in lsearch files need not
94 contain colons. This was done some time before quoting was made available
95 in lsearch files. However, iplsearch files do require colons in IPv6 keys
96 (notated using the quote facility) so as to distinguish them from IPv4
97 keys. This meant that lookups for IP addresses in host lists did not work
98 for iplsearch lookups.
100 This has been fixed by arranging for IPv6 addresses to be expressed with
101 colons if the lookup type is iplsearch. This is not incompatible, because
102 previously such lookups could never work.
104 The situation is now rather anomolous, since one *can* have colons in
105 ordinary lsearch keys. However, making the change in all cases is
106 incompatible and would probably break a number of configurations.
108 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
111 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
112 conversion specification without a maximum field width, thereby enabling
113 a rogue spamd server to cause a buffer overflow. While nobody in their
114 right mind would setup Exim to query an untrusted spamd server, an
115 attacker that gains access to a server running spamd could potentially
116 exploit this vulnerability to run arbitrary code as the Exim user.
118 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
119 $primary_hostname instead of what libspf2 thinks the hosts name is.
121 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
122 a directory entry by the name of the lookup key. Previously, if a
123 symlink pointed to a non-existing file or a file in a directory that
124 Exim lacked permissions to read, a lookup for a key matching that
125 symlink would fail. Now it is enough that a matching directory entry
126 exists, symlink or not. (Bugzilla 503.)
128 PH/03 The body_linecount and body_zerocount variables are now exported in the
131 PH/04 Added the $dnslist_matched variable.
133 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
134 This means they are set thereafter only if the connection becomes
137 PH/06 Added the client_condition to authenticators so that some can be skipped
138 by clients under certain conditions.
140 PH/07 The error message for a badly-placed control=no_multiline_responses left
141 "_responses" off the end of the name.
143 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
145 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
146 (without spaces) instead of just copying the configuration text.
148 PH/10 Added the /noupdate option to the ratelimit ACL condition.
150 PH/11 Added $max_received_linelength.
152 PH/12 Added +ignore_defer and +include_defer to host lists.
154 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
155 way in which PCRE > 7.0 is built.
157 PH/14 Implemented queue_only_load_latch.
159 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
160 MAIL command. The effect was to mangle the value on 64-bit systems.
162 PH/16 Another patch from the Sieve maintainer.
164 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
166 PH/18 If a system quota error occurred while trying to create the file for
167 a maildir delivery, the message "Mailbox is full" was not appended to the
168 bounce if the delivery eventually timed out. Change 4.67/27 below applied
169 only to a quota excession during the actual writing of the file.
171 PH/19 It seems that peer DN values may contain newlines (and other non-printing
172 characters?) which causes problems in log lines. The DN values are now
173 passed through string_printing() before being added to log lines.
175 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
176 and InterBase are left for another time.)
178 PH/21 Added message_body_newlines option.
180 PH/22 Guard against possible overflow in moan_check_errorcopy().
182 PH/23 POSIX allows open() to be a macro; guard against that.
184 PH/24 If the recipient of an error message contained an @ in the local part
185 (suitably quoted, of course), incorrect values were put in $domain and
186 $local_part during the evaluation of errors_copy.
192 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
193 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
196 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
197 issue a MAIL command.
199 PH/02 In an ACL statement such as
201 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
203 if a client was not listed at all, or was listed with a value other than
204 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
205 the condition was not true (as it should be), so access was not denied.
206 The bug was that the ! inversion was incorrectly passed on to the second
207 item. This has been fixed.
209 PH/03 Added additional dnslists conditions == and =& which are different from
210 = and & when the dns lookup returns more than one IP address.
212 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
213 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
215 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
216 FSYNC, which compiles an option called disable_fsync that allows for
217 bypassing fsync(). The documentation is heavily laced with warnings.
219 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
221 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
222 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
223 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
224 including adding "make clean"; (3) Added -fPIC when compiling the test
225 dynamically loaded module, to get rid of a warning.
227 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
228 message fails, move_frozen_messages = true and ignore_bounce_errors_after
229 = 0s. The bug is otherwise harmless.
231 PH/07 There was a bug in the dovecot authenticator such that the value of
232 $auth1 could be overwritten, and so not correctly preserved, after a
233 successful authentication. This usually meant that the value preserved by
234 the server_setid option was incorrect.
236 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
238 PH/09 Installed PCRE release 7.0.
240 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
241 run for batched SMTP input. It is now run at the start of every message
242 in the batch. While fixing this I discovered that the process information
243 (output by running exiwhat) was not always getting set for -bs and -bS
244 input. This is fixed, and it now also says "batched" for BSMTP.
246 PH/11 Added control=no_pipelining.
248 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
249 patch, slightly modified), and move the expansion of helo_data till after
250 the connection is made in the smtp transport (so it can use these
253 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
255 PH/14 Added log_selector = +pid.
257 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
259 PH/16 Add ${if forany and ${if forall.
261 PH/17 Added dsn_from option to vary the From: line in DSNs.
263 PH/18 Flush SMTP output before performing a callout, unless control =
264 no_callout_flush is set.
266 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
267 was true (the default) a successful delivery failed to delete the retry
268 item, thus causing premature timeout of the address. The bug is now
271 PH/20 Added hosts_avoid_pipelining to the smtp transport.
273 PH/21 Long custom messages for fakedefer and fakereject are now split up
274 into multiline reponses in the same way that messages for "deny" and
275 other ACL rejections are.
277 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
278 with slight modification.
280 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
281 draft, changing the syntax and factoring some duplicate code".
283 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
284 for deliveries of the second and subsequent messages over the same SMTP
287 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
288 ${reduce, with only minor "tidies".
290 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
292 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
293 expansion side effects.
295 PH/27 When a message times out after an over-quota error from an Exim-imposed
296 quota, the bounce message says "mailbox is full". This message was not
297 being given when it was a system quota that was exceeded. It now should
300 MH/03 Made $recipients available in local_scan(). local_scan() already has
301 better access to the recipient list through recipients_list[], but
302 $recipients can be useful in postmaster-provided expansion strings.
304 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
305 in the case of a MAIL command with additional options following the
306 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
307 were accidentally chopped off.
309 PH/29 SMTP synchronization checks are implemented when a command is read -
310 there is a check that no more input is waiting when there shouldn't be
311 any. However, for some commands, a delay in an ACL can mean that it is
312 some time before the response is written. In this time, more input might
313 arrive, invalidly. So now there are extra checks after an ACL has run for
314 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
315 pipelining has not been advertised.
317 PH/30 MH's patch to allow iscntrl() characters to be list separators.
319 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
320 returned in the SMTP response when smtp_return_error_details was false.
323 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
324 instead of the C I/O that was originally supplied, because problems were
327 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
328 Exim which did not show up earlier: it was assuming that a call to
329 SSL_CTX_set_info_callback() might give an error value. In fact, there is
330 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
331 was a macro that became an assignment, so it seemed to work. This has
332 changed to a proper function call with a void return, hence the compile
333 error. Exim's code has been fixed.
335 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
338 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
340 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
342 PH/37 If a message is not accepted after it has had an id assigned (e.g.
343 because it turns out to be too big or there is a timeout) there is no
344 "Completed" line in the log. When some messages of this type were
345 selected by exigrep, they were listed as "not completed". Others were
346 picked up by some special patterns. I have improved the selection
347 criteria to be more general.
349 PH/38 The host_find_failed option in the manualroute router can now be set
350 to "ignore", to completely ignore a host whose IP address cannot be
351 found. If all hosts are ignored, the behaviour is controlled by the new
352 host_all_ignored option.
354 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
355 homing or because of multiple MX records with /mx) generated more than
356 one IP address, and the following item turned out to be the local host,
357 all the secondary addresses of the first item were incorrectly removed
358 from the list, along with the local host and any following hosts (which
359 is what is supposed to happen).
361 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
362 whoever called Exim into the -H file. In the case of the daemon it was
363 behaving confusingly. When first started, it used values for whoever
364 started the daemon, but after a SIGHUP it used the Exim user (because it
365 calls itself on a restart). I have changed the code so that it now always
368 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
369 message are rejected with the same error (e.g. no authentication or bad
370 sender address), and a DATA command is nevertheless sent (as can happen
371 with PIPELINING or a stupid MUA), the error message that was given to the
372 RCPT commands is included in the rejection of the DATA command. This is
373 intended to be helpful for MUAs that show only the final error to their
376 PH/42 Another patch from the Sieve maintainer.
378 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
379 Eximstats - Fixed some broken HTML links and added missing column headers
381 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
382 columns for Rejects, Temp Rejects, Ham, and Spam rows.
384 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
386 PH/43 Yet another patch from the Sieve maintainer.
388 PH/44 I found a way to check for a TCP/IP connection going away before sending
389 the response to the final '.' that terminates a message, but only in the
390 case where the client has not sent further data following the '.'
391 (unfortunately, this is allowed). However, in many cases there won't be
392 any further data because there won't be any more messages to send. A call
393 to select() can be used: if it shows that the input is "ready", there is
394 either input waiting, or the socket has been closed. An attempt to read
395 the next input character can distinguish the two cases. Previously, Exim
396 would have sent an OK response which the client would never have see.
397 This could lead to message repetition. This fix should cure that, at
398 least in a lot of common cases.
400 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
401 advertised in response to EHLO.
407 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
408 fixed by 4.65/MH/01 (is this a record?) are fixed:
410 (i) An empty string was always treated as zero by the numeric comparison
411 operators. This behaviour has been restored.
413 (ii) It is documented that the numeric comparison operators always treat
414 their arguments as decimal numbers. This was broken in that numbers
415 starting with 0 were being interpreted as octal.
417 While fixing these problems I realized that there was another issue that
418 hadn't been noticed. Values of message_size_limit (both the global option
419 and the transport option) were treated as octal if they started with 0.
420 The documentation was vague. These values are now always treated as
421 decimal, and I will make that clear in the documentation.
427 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
428 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
431 MH/01 Don't check that the operands of numeric comparison operators are
432 integers when their expansion is in "skipping" mode (fixes bug
433 introduced by 4.64-PH/07).
435 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
436 child addresses, Exim now panics and dies. Previously, because the count
437 is held in a short int, deliveries were likely to be lost. As such a
438 large number of recipients for a single message is ridiculous
439 (performance will be very, very poor), I have chosen to impose a limit
440 rather than extend the field.
446 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
447 leftover -K file (the existence of which was triggered by #402).
448 While we were at it, introduced process PID as part of the -K
449 filename. This should rule out race conditions when creating
452 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
453 processing considerably. Previous code took too long for large mails,
454 triggering a timeout which in turn triggers #401.
456 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
457 in the DK code in transports.c. sendfile() is not really portable,
458 hence the _LINUX specificness.
460 TF/01 In the add_headers option to the mail command in an Exim filter,
461 there was a bug that Exim would claim a syntax error in any
462 header after the first one which had an odd number of characters
465 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
466 callout verification, Exim cached a "reject" for the entire domain. This
467 is correct for most verifications, but it is not correct for a recipient
468 verification with use_sender or use_postmaster set, because in that case
469 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
470 case of MAIL FROM:<> rejection from other early rejections (e.g.
471 rejection of HELO). When verifying a recipient using a non-null MAIL
472 address, the cache is ignored if it shows MAIL FROM:<> rejection.
473 Whatever the result of the callout, the value of the domain cache is
474 left unchanged (for any other kind of callout, getting as far as trying
475 RCPT means that the domain itself is ok).
477 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
480 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
481 manifest itself as EPIPE rather than ECONNECT. When tidying away a
482 session, the daemon ignores ECONNECT errors and logs others; it now
483 ignores EPIPE as well.
485 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
486 (quoted-printable decoding).
488 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
489 later the small subsequent patch to fix an introduced bug.
491 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
493 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
495 PH/08 An error is now given if message_size_limit is specified negative.
497 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
498 to be given (somewhat) arbitrary names.
500 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
503 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
504 miscellaneous code fixes
506 PH/10 Added the log_reject_target ACL modifier to specify where to log
509 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
510 hostname. This is wrong, because it relates to the incoming message (and
511 probably the interface on which it is arriving) and not to the outgoing
512 callout (which could be using a different interface). This has been
513 changed to use the value of the helo_data option from the smtp transport
514 instead - this is what is used when a message is actually being sent. If
515 there is no remote transport (possible with a router that sets up host
516 addresses), $smtp_active_hostname is used.
518 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
519 tweaks were necessary in order to get it to work (see also 21 below):
520 (a) The code assumed that strncpy() returns a negative number on buffer
521 overflow, which isn't the case. Replaced with Exim's string_format()
523 (b) There were several signed/unsigned issues. I just did the minimum
524 hacking in of casts. There is scope for a larger refactoring.
525 (c) The code used strcasecmp() which is not a standard C function.
526 Replaced with Exim's strcmpic() function.
527 (d) The code set only $1; it now sets $auth1 as well.
528 (e) A simple test gave the error "authentication client didn't specify
529 service in request". It would seem that Dovecot has changed its
530 interface. Fortunately there's a specification; I followed it and
531 changed what the client sends and it appears to be working now.
533 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
536 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
537 address A is aliased to B and C, where B exists and C does not. Without
538 -v the output is "A verified" because verification stops after a
539 successful redirection if more than one address is generated. However,
540 with -v the child addresses are also verified. Exim was outputting "A
541 failed to verify" and then showing the successful verification for C,
542 with its parentage. It now outputs "B failed to verify", showing B's
543 parentage before showing the successful verification of C.
545 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
546 look up a TXT record in a specific list after matching in a combined
549 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
550 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
551 they consult the DNS. I had assumed they would set it the way they
552 wanted; and indeed my experiments on Linux seem to show that in some
553 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
554 To be on the safe side, however, I have now made the interface to
555 host_find_byname() similar to host_find_bydns(), with an argument
556 containing the DNS resolver options. The host_find_byname() function now
557 sets these options at its start, just as host_find_bydns() does. The smtp
558 transport options dns_qualify_single and dns_search_parents are passed to
559 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
560 of host_find_byname() use the default settings of RES_DEFNAMES
561 (qualify_single) but not RES_DNSRCH (search_parents).
563 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
564 spool_read_header() do less string testing, by means of a preliminary
565 switch on the second character of optional "-foo" lines. (This is
566 overdue, caused by the large number of possibilities that now exist.
567 Originally there were few.) While I was there, I also converted the
568 str(n)cmp tests so they don't re-test the leading "-" and the first
569 character, in the hope this might squeeze out yet more improvement.
571 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
572 flag allowing group syntax was set by the header_syntax check but not
573 turned off, possible causing trouble later; (2) The flag was not being
574 set at all for the header_verify test, causing "group"-style headers to
575 be rejected. I have now set it in this case, and also caused header_
576 verify to ignore an empty address taken from a group. While doing this, I
577 came across some other cases where the code for allowing group syntax
578 while scanning a header line wasn't quite right (mostly, not resetting
579 the flag correctly in the right place). These bugs could have caused
580 trouble for malformed header lines. I hope it is now all correct.
582 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
583 with the "reply" argument non-NULL. The code, however (which originally
584 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
585 but it didn't always do it. This confused somebody who was copying the
586 code for some other use. I have removed all the tests.
588 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
589 feature that was used to support insecure browsers during the U.S. crypto
590 embargo. It requires special client support, and Exim is probably the
591 only MTA that supported it -- and would never use it because real RSA is
592 always available. This code has been removed, because it had the bad
593 effect of slowing Exim down by computing (never used) parameters for the
594 RSA_EXPORT functionality.
596 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
597 authenticator to fail if there's a tab character in the incoming data
598 (there should never be unless someone is messing about, as it's supposed
599 to be base64-encoded). Also added, on Timo's advice, the "secured" option
600 if the connection is using TLS or if the remote IP is the same as the
601 local IP, and the "valid-client-cert option" if a client certificate has
604 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
605 authenticators. This can be used for authorization after authentication
606 succeeds. (In the case of plaintext, it servers for both authentication
609 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
610 if any retry times were supplied.
612 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
613 connection, where there is no client IP address to check. In this
614 situation, the verify now always succeeds.
616 PH/25 Applied John Jetmore's -Mset patch.
618 PH/26 Added -bem to be like -Mset, but loading a message from a file.
620 PH/27 In a string expansion for a processed (not raw) header when multiple
621 headers of the same name were present, leading whitespace was being
622 removed from all of them, but trailing whitespace was being removed only
623 from the last one. Now trailing whitespace is removed from each header
624 before concatenation. Completely empty headers in a concatenation (as
627 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
628 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
630 PH/29 [Removed. This was a change that I later backed out, and forgot to
631 correct the ChangeLog entry (that I had efficiently created) before
632 committing the later change.]
634 PH/30 Exim was sometimes attempting to deliver messages that had suffered
635 address errors (4xx response to RCPT) over the same connection as other
636 messages routed to the same hosts. Such deliveries are always "forced",
637 so retry times are not inspected. This resulted in far too many retries
638 for the affected addresses. The effect occurred only when there were more
639 hosts than the hosts_max_try setting in the smtp transport when it had
640 the 4xx errors. Those hosts that it had tried were not added to the list
641 of hosts for which the message was waiting, so if all were tried, there
642 was no problem. Two fixes have been applied:
644 (i) If there are any address or message errors in an SMTP delivery, none
645 of the hosts (tried or untried) are now added to the list of hosts
646 for which the message is waiting, so the message should not be a
647 candidate for sending over the same connection that was used for a
648 successful delivery of some other message. This seems entirely
649 reasonable: after all the message is NOT "waiting for some host".
650 This is so "obvious" that I'm not sure why it wasn't done
651 previously. Hope I haven't missed anything, but it can't do any
652 harm, as the worst effect is to miss an optimization.
654 (ii) If, despite (i), such a delivery is accidentally attempted, the
655 routing retry time is respected, so at least it doesn't keep
656 hammering the server.
658 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
659 in ${readsocket because some servers need this prod.
661 PH/32 Added some extra debug output when updating a wait-xxx database.
663 PH/33 The hint "could be header name not terminated by colon", which has been
664 given for certain expansion errors for a long time, was not being given
665 for the ${if def:h_colon_omitted{... case.
667 PH/34 The spec says: "With one important exception, whenever a domain list is
668 being scanned, $domain contains the subject domain." There was at least
669 one case where this was not true.
671 PH/35 The error "getsockname() failed: connection reset by peer" was being
672 written to the panic log as well as the main log, but it isn't really
673 panic-worthy as it just means the connection died rather early on. I have
674 removed the panic log writing for the ECONNRESET error when getsockname()
677 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
678 runs only) independently of the message's sender address. This meant
679 that, if the 4xx error was in fact related to the sender, a different
680 message to the same recipient with a different sender could confuse
681 things. In particualar, this can happen when sending to a greylisting
682 server, but other circumstances could also provoke similar problems.
683 I have changed the default so that the retry time for these errors is now
684 based a combination of the sender and recipient addresses. This change
685 can be overridden by setting address_retry_include_sender=false in the
688 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
689 remote server are returned as part of bounce messages. This was not
690 happening for LMTP over a pipe (the lmtp transport), but now it is the
691 same for both kinds of LMTP.
693 PH/38 Despite being documented as not happening, Exim was rewriting addresses
694 in header lines that were in fact CNAMEs. This is no longer the case.
696 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
697 and queue runs started by the daemon processed all messages. This has
698 been fixed so that -R and -S can now usefully be given with -q<time>.
700 PH/40 Import PCRE release 6.7 (fixes some bugs).
702 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
704 PH/42 Give an error if -q is specified more than once.
706 PH/43 Renamed the variables $interface_address and $interface_port as
707 $received_ip_address and $received_port, to make it clear that these
708 values apply to message reception, and not to the outgoing interface when
709 a message is delivered. (The old names remain recognized, of course.)
711 PH/44 There was no timeout on the connect() call when using a Unix domain
712 socket in the ${readsocket expansion. There now is.
714 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
715 be meaningful with "accept".
717 SC/01 Eximstats V1.43
718 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
720 SC/02 Eximstats V1.44
721 Use a glob alias rather than an array ref in the generated
722 parser. This improves both readability and performance.
724 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
725 Collect SpamAssassin and rejection statistics.
726 Don't display local sender or destination tables unless
727 there is data to show.
728 Added average volumes into the top table text output.
730 SC/04 Eximstats V1.46
731 Collect data on the number of addresses (recipients)
732 as well as the number of messages.
734 SC/05 Eximstats V1.47
735 Added 'Message too big' to the list of mail rejection
736 reasons (thanks to Marco Gaiarin).
738 SC/06 Eximstats V1.48
739 Mainlog lines which have GMT offsets and are too short to
740 have a flag are now skipped.
742 SC/07 Eximstats V1.49 (Alain Williams)
743 Added the -emptyok flag.
745 SC/08 Eximstats V1.50
746 Fixes for obtaining the IP address from reject messages.
748 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
749 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
750 whitesspace changes from 4.64-PH/27
752 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
755 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
756 are found, allow negative numbers in numeric criteria)
758 JJ/06 exipick.20061117.2, added new $message_body_missing variable
760 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
761 to match changes made in 4.64-PH/43
763 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
765 PH/47 Put in an explicit test for a DNS lookup of an address record where the
766 "domain" is actually an IP address, and force a failure. This locks out
767 those revolvers/nameservers that support "A-for-A" lookups, in
768 contravention of the specifications.
770 PH/48 When a host name was looked up from an IP address, and the subsequent
771 forward lookup of the name timed out, the host name was left in
772 $sender_host_name, contrary to the specification.
774 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
775 restricted to single-key lookups, Exim was not diagnosing an error if
776 * or *@ was used with a query-style lookup.
778 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
780 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
781 long ago, but noone interested enough thought of it. Let's just say that
782 the "1.1" means that there are some new functions that weren't there at
783 some point in the past.
785 PH/51 Error processing for expansion failure of helo_data from an smtp
786 transport during callout processing was broken.
788 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
789 tested/used via the -bh/-bhc/-bs options.
791 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
792 bug, fixed in subsequent PCRE releases).
794 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
795 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
797 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
803 SC/01 Use a glob alias rather than an array ref in eximstats generated
804 parser. This improves both readability and performance.
806 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
807 Don't display local sender or destination tables in eximstats unless
808 there is data to show.
809 Added average volumes into the eximstats top table text output.
811 SC/03 Collect data on the number of addresses (recipients) as well
812 as the number of messages in eximstats.
814 TF/01 Correct an error in the documentation for the redirect router. Exim
815 does (usually) call initgroups() when daemonizing.
817 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
818 with consistent privilege compared to when running as a daemon.
820 TF/03 Note in the spec that $authenticated_id is not set for local
821 submissions from trusted users.
823 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
824 Thanks to Dean Brooks <dean@iglou.com> for the patch.
826 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
827 by adding some example configuration directives to the default
828 configuration file. A little bit of work is required to uncomment the
829 directives and define how usernames and passwords are checked, but
830 there is now a framework to start from.
832 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
833 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
834 without this. I don't know how relevant this is to other LDAP libraries.
836 PH/02 Add the verb name to the "unknown ACL verb" error.
838 PH/03 Magnus Holmgren's patch for filter_prepend_home.
840 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
842 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
843 directory not expanded when it should be if an expanded home directory
844 was set for the address (which is overridden by the transport).
846 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
849 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
850 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
851 because it is too late at that time, and has no effect.
853 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
854 security issue with \' (bugzilla #107). I could not use the
855 PQescapeStringConn() function, because it needs a PGconn value as one of
858 PH/08 When testing addresses using -bt, indicate those final addresses that
859 are duplicates that would not cause an additional delivery. At least one
860 person was confused, thinking that -bt output corresponded to deliveries.
861 (Suppressing duplicates isn't a good idea as you lose the information
862 about possibly different redirections that led to the duplicates.)
864 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
865 systems where poll() doesn't work, in particular OS X.
867 PH/10 Added more information to debugging output for retry time not reached.
869 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
870 operations in malware.c.
872 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
875 PH/13 If write_rejectlog was set false when logging was sent to syslog with
876 syslog_duplication set false, log lines that would normally be written
877 both the the main log and to the reject log were not written to syslog at
880 PH/14 In the default configuration, change the use of "message" in ACL warn
881 statements to "add_header".
883 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
884 not followed by a command (e.g. "seen endif").
886 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
887 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
890 PH/17 Added extra conditions to the default value of delay_warning_condition
894 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
895 { match{$h_precedence:}{(?i)bulk|list|junk} } \
896 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
899 The Auto-Submitted: and various List- headers are standardised, whereas I
900 don't think Precedence: ever was.
902 PH/18 Refactored debugging code in route_finduser() to show more information,
903 in particular, the error code if getpwnam() issues one.
905 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
906 This is apparently needed in addition to the PH/07 change above to avoid
907 any possible encoding problems.
909 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
910 but not after initializing Perl.
912 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
913 output them only if debugging. By default they are written stderr,
914 apparently, which is not desirable.
916 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
919 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
922 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
924 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
925 authenticated or an ident call has been made. Suppress the default
926 values for $authenticated_id and $authenticated_sender (but permit -oMai
927 and -oMas) when testing with -bh.
929 PH/24 Re-jigged the order of the tests in the default configuration so that the
930 tests for valid domains and recipients precede the DNS black list and CSA
931 tests, on the grounds that those ones are more expensive.
933 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
934 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
935 This bug exists in every version of Exim that I still have, right back to
938 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
939 However, an attempt to turn on case-sensitivity in a regex key by
940 including (?-i) didn't work because the subject string was already
941 lowercased, and the effects were non-intuitive. It turns out that a
942 one-line patch can be used to allow (?-i) to work as expected.
948 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
949 other effects) broke the use of negated acl sub-conditions.
951 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
954 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
955 "Deny" causes Exim to reject the incoming connection with a 554 error.
956 Unfortunately, if there is a major crisis, such as a disk failure,
957 tcp-wrappers gives "deny", whereas what one would like would be some
958 kind of temporary error. A kludge has been added to help with this.
959 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
960 554 error is used if errno is still zero or contains ENOENT (which occurs
961 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
964 PH/03 Add -lutil to the default FreeBSD LIBS setting.
966 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
967 errors. Otherwise a message that provokes a temporary error (when other
968 messages do not) can cause a whole host to time out.
970 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
971 addresses were routed directly to files or pipes from a redirect router.
972 File deliveries just didn't batch; pipe deliveries might have suffered
975 PH/06 A failure to get a lock for a hints database would erroneously always say
976 "Failed to get write lock", even when it was really a read lock.
978 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
979 of 0600. This has been changed to use the value of the lockfile_mode
980 option (which defaults to 0600).
982 PH/08 Applied small patch from the Sieve maintainer.
984 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
985 folder from quota calculations, a direct delivery into this folder messed
986 up the contents of the maildirsize file. This was because the regex was
987 used only to exclude .Trash (or whatever) when the size of the mailbox
988 was calculated. There was no check that a delivery was happening into an
989 excluded directory. This bug has been fixed by ignoring all quota
990 processing for deliveries into excluded directories.
992 PH/10 Added the maildirfolder_create_regex option to appendfile.
998 PH/01 The code for finding all the local interface addresses on a FreeBSD
999 system running IPv6 was broken. This may well have applied to all BSD
1000 systems, as well as to others that have similar system calls. The broken
1001 code found IPv4 interfaces correctly, but gave incorrect values for the
1002 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1003 that it would not match correctly against @[] and not recognize the IPv6
1006 PH/02 The ipliteral router was not recognizing addresses of the form user@
1007 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1009 PH/03 Added disable_ipv6.
1011 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1012 decoded form, because it is most often used to construct To: headers
1013 lines in autoreplies, and the decoded form may well be syntactically
1014 invalid. However, $reply_address has leading white space removed, and all
1015 newlines turned into spaces so that the autoreply transport does not
1018 PH/05 If group was specified without a user on a router, and no group or user
1019 was specified on a transport, the group from the router was ignored.
1021 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1022 for visible compile-time settings that can be used to change these
1023 numbers, for those that want even more. Backwards compatibility with old
1024 spool files has been maintained. However, going back to a previous Exim
1025 release will lost any variables that are in spool files.
1027 PH/07 Two small changes when running in the test harness: increase delay when
1028 passing a TCP/IP connection to a new process, in case the original
1029 process has to generate a bounce, and remove special handling of
1030 127.0.0.2 (sic), which is no longer necessary.
1032 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1033 be the same on different OS.
1035 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1038 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1039 whether --show-vars was specified or not
1041 JJ/02 exipick: Added support for new ACL variable spool format introduced
1044 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1045 syntactically invalid From: or Reply-to: line, and a filter used this to
1046 generate an autoreply, and therefore failed to obtain an address for the
1047 autoreply, Exim could try to deliver to a non-existent relative file
1048 name, causing unrelated and misleading errors. What now happens is that
1049 it logs this as a hard delivery error, but does not attempt to create a
1052 PH/11 The exinext utility has a -C option for testing purposes, but although
1053 the given file was scanned by exinext itself; it wasn't being passed on
1054 when Exim was called.
1056 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1057 an end-of-file indication when reading a command response.
1059 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1060 compiled. In many other places in Exim, IPv6 addresses are always
1061 recognized, so I have changed this. It also means that IPv4 domain
1062 literals of the form [IPV4:n.n.n.n] are now always recognized.
1064 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1065 used if the router is not running as root, for example, when verifying at
1066 ACL time, or when using -bh. The debugging output from this situation was
1067 non-existent - all you got was a failure to exec. I have made two
1070 (a) Failures to set uid/gid, the current directory, or a process leader
1071 in a subprocess such as that created by queryprogram now generate
1072 suitable debugging ouput when -d is set.
1074 (b) The queryprogram router detects when it is not running as root,
1075 outputs suitable debugging information if -d is set, and then runs
1076 the subprocess without attempting to change uid/gid.
1078 PH/15 Minor change to Makefile for building test_host (undocumented testing
1081 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1082 additional section of a DNS packet that returns MX or SRV records.
1083 Instead, it always explicitly searches for A/AAAA records. This avoids
1084 major problems that occur when a DNS server includes only records of one
1085 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1086 fixed another bug: if SRV records were looked up and the corresponding
1087 address records were *not* found in the additional section, the port
1088 values from the SRV records were lost.
1090 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1091 using the correct key (the original address) when searching the retry
1092 rules in order to find which one to use for generating the retry hint.
1094 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1095 adding the default one. Similarly, if it contains a Reply-To: header, the
1096 errors_reply_to option, if set, is not used.
1098 PH/19 When calculating a retry time, Exim used to measure the "time since
1099 failure" by looking at the "first failed" field in the retry record. Now
1100 it does not use this if it is later than than the arrival time of the
1101 message. Instead it uses the arrival time. This makes for better
1102 behaviour in cases where some deliveries succeed, thus re-setting the
1103 "first failed" field. An example is a quota failure for a huge message
1104 when small messages continue to be delivered. Without this change, the
1105 "time since failure" will always be short, possible causing more frequent
1106 delivery attempts for the huge message than are intended.
1107 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1109 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1110 $1, $2, $3) because the numerical variables can be reset during some
1111 expansion items (e.g. "match"), thereby losing the authentication data.
1113 PH/21 Make -bV show the size of off_t variables so that the test suite can
1114 decide whether to run tests for quotas > 2G.
1116 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1117 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1118 filecount value is greater than 2G or if a quota value is greater than 2G
1119 on a system where the size of off_t is not greater than 4, a panic error
1122 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1123 never match. The debug and -bh output now contains an explicit error
1124 message indicating a malformed IPv4 address or mask.
1126 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1127 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1128 PH/23 above applies.
1130 PH/25 Do not write to syslog when running in the test harness. The only
1131 occasion when this arises is a failure to open the main or panic logs
1132 (for which there is an explicit test).
1134 PH/26 Added the /no_tell option to "control=freeze".
1136 PH/27 If a host name lookup failed very early in a connection, for example, if
1137 the IP address matched host_lookup and the reverse lookup yielded a name
1138 that did not have a forward lookup, an error message of the form "no IP
1139 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1140 could be logged. Now it outputs the IP address instead of "NULL".
1142 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1143 allows the sender and the authenticated sender to be set when
1144 submitting a message from within Exim. Since child_open_exim() is
1145 documented for local_scan(), the new function should be too.
1147 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1148 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1149 results in an empty string is now treated as unset.
1151 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1153 PH/31 Added sender_verify_fail logging option.
1155 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1156 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1157 tidied the source and removed it altogether.
1159 PH/33 When a queue run was abandoned because the load average was too high, a
1160 log line was always written; now it is written only if the queue_run log
1161 selector is set. In addition, the log line for abandonment now contains
1162 information about the queue run such as the pid. This is always present
1163 in "start" and "stop" lines but was omitted from the "abandon" line.
1165 PH/34 Omit spaces between a header name and the colon in the error message that
1166 is given when verify = headers_syntax fails (if there are lots of them,
1167 the message gets confusing).
1169 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1170 names, as there are now some PTR records that contain slashes. This check
1171 is only to protect against broken name servers that fall over on strange
1172 characters, so the fact that it applies to all lookups doesn't matter.
1174 PH/36 Now that the new test suite is complete, we can remove some of the
1175 special code in Exim that was needed for the old test suite. For example,
1176 sorting DNS records because real resolvers return them in an arbitrary
1177 order. The new test suite's fake resolver always returns records in the
1180 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1181 bounces) except when queue_only is set, to avoid logging races between
1182 the different processes.
1184 PH/38 Panic-die if .include specifies a non-absolute path.
1186 PH/39 A tweak to the "H" retry rule from its user.
1188 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1189 a label. They prevented compilation on older perls.
1191 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1192 a warning to be raised on newish perls.
1194 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1195 on queue. Changes to match documented behaviour of showing count of
1196 messages matching specified criteria.
1198 PH/40 Changed the default ident timeout from 30s to 5s.
1200 PH/41 Added support for the use of login_cap features, on those BSD systems
1201 that have them, for controlling the resources used by pipe deliveries.
1203 PH/42 The content-scanning code uses fopen() to create files in which to put
1204 message data. Previously it was not paying any attention to the mode of
1205 the files. Exim runs with umask(0) because the rest of the code creates
1206 files with open(), and sets the required mode explicitly. Thus, these
1207 files were ending up world-writeable. This was not a big issue, because,
1208 being within the spool directory, they were not world-accessible. I have
1209 created a function called modefopen, which takes an additional mode
1210 argument. It sets umask(777), creates the file, chmods it to the required
1211 mode, then resets the umask. All the relevant calls to fopen() in the
1212 content scanning code have been changed to use this function.
1214 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1215 to 24 hours. This avoids potential overflow problems when processing G
1216 and H retry rules. I suspect nobody ever tinkers with this value.
1218 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1220 PH/45 When the plaintext authenticator is running as a client, the server's
1221 challenges are checked to ensure they are valid base64 strings. By
1222 default, the authentication attempt is cancelled if an invalid string is
1223 received. Setting client_ignore_invalid_base64 true ignores these errors.
1224 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1225 they are received. Thus, the responses can be made to depend on the
1226 challenges. If an invalid string is ignored, an empty string is placed in
1229 PH/46 Messages that are created by the autoreply transport now contains a
1230 References: header, in accordance with RFCs 2822 and 3834.
1232 PH/47 Added authenticated_sender_force to the smtp transport.
1234 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1236 PH/49 Installed latest patch from the Sieve maintainer.
1238 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1239 was also set, the appendfile transport was unnecessarily scanning a
1240 directory of message files (e.g. for maildir delivery) to find the count
1241 of files (along with the size), even though it did not need this
1242 information. It now does the scan only if it needs to find either the
1243 size of the count of files.
1245 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1247 PH/52 Two bugs concerned with error handling when the smtp transport is
1250 (i) Exim was not creating retry information for temporary errors given
1251 for individual recipients after the DATA command when the smtp transport
1252 was used in LMTP mode. This meant that they could be retried too
1253 frequently, and not timed out correctly.
1255 (ii) Exim was setting the flag that allows error details to be returned
1256 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1257 recipients that were returned after the DATA command.
1259 PH/53 This is related to PH/52, but is more general: for any failing address,
1260 when detailed error information was permitted to be returned to the
1261 sender, but the error was temporary, then after the final timeout, only
1262 "retry timeout exceeded" was returned. Now it returns the full error as
1263 well as "retry timeout exceeded".
1265 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1266 do this, and (what is worse) MTAs that accept it.
1268 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1269 will now be deprecated.
1271 PH/56 New os.c-cygwin from the Cygwin maintainer.
1273 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1274 formats (previously only available in exim formats via -bpr, -bpru,
1275 and -bpra. Now also available in native and exiqgrep formats)
1277 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1278 with very large, slow to parse queues
1280 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1282 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1284 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1285 responses to authentication challenges, though it was showing the
1286 challenges; (ii) I've removed the CR characters from the debug output for
1289 PH/58 Allow for the insertion of a newline as well as a space when a string
1290 is turned into more than one encoded-word during RFC 2047 encoding. The
1291 Sieve code now uses this.
1293 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1294 data_4xx, lost_connection, tls_required.
1296 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1297 message was being sent as an SMTP response.
1299 PH/61 Add -l and -k options to exicyclog.
1301 PH/62 When verifying, if an address was redirected to one new address, so that
1302 verification continued, and the new address failed or deferred after
1303 having set something in $address_data, the value of $address_data was not
1304 passed back to the ACL. This was different to the case when no
1305 redirection occurred. The value is now passed back in both cases.
1307 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1308 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1309 use login_cap.h, so on its own it isn't the distinguishing feature. The
1310 new name refers directly to the setclassresources() function.
1312 PH/65 Added configuration files for NetBSD3.
1314 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1316 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1317 is preferred over IPv4.
1319 PH/68 The bounce_return_message and bounce_return_body options were not being
1320 honoured for bounces generated during the reception of non-SMTP messages.
1321 In particular, this applied to messages rejected by the ACL. This bug has
1322 been fixed. However, if bounce_return_message is true and bounce_return_
1323 body is false, the headers that are returned for a non-SMTP message
1324 include only those that have been read before the error was detected.
1325 (In the case of an ACL rejection, they have all been read.)
1327 PH/69 The HTML version of the specification is now built in a directory called
1328 spec_html instead of spec.html, because the latter looks like a path with
1329 a MIME-type, and this confuses some software.
1331 PH/70 Catch two compiler warnings in sieve.c.
1333 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1334 function verify_get_ident() calls ip_connect() to connect a socket, but
1335 if the "connect()" function timed out, ip_connect() used to close the
1336 socket. However, verify_get_ident() also closes the socket later, and in
1337 between Exim writes to the log, which may get opened at this point. When
1338 the socket was closed in ip_connect(), the log could get the same file
1339 descriptor number as the socket. This naturally causes chaos. The fix is
1340 not to close the socket in ip_connect(); the socket should be closed by
1341 the function that creates it. There was only one place in the code where
1342 this was missing, in the iplookup router, which I don't think anybody now
1343 uses, but I've fixed it anyway.
1345 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1346 well as to direct DNS lookups. Otherwise the handling of names in host
1347 lists is inconsistent and therefore confusing.
1353 PH/01 Two changes to the default runtime configuration:
1355 (1) Move the checks for relay_from_hosts and authenticated clients from
1356 after to before the (commented out) DNS black list checks.
1358 (2) Add control=submission to the relay_from_hosts and authenticated
1359 clients checks, on the grounds that messages accepted by these
1360 statements are most likely to be submissions.
1362 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1364 (1) Generate an error if the third argument for the ${prvs expansion is
1367 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1370 (3) Reset the variables that are obtained from the first argument of
1371 ${prvscheck and used in the second argument before leaving the code,
1372 because their memory is reclaimed, so using them afterwards may do
1375 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1376 one (it's much easier than Tom thought :-).
1378 (5) Because of (4), we can now allow for the use of $prvscheck_result
1379 inside the third argument.
1381 PH/03 For some reason, the default setting of PATH when running a command from
1382 a pipe transport was just "/usr/bin". I have changed it to
1385 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1386 anything to be listed in the output from -bV.
1388 PH/05 When a filter generated an autoreply, the entire To: header line was
1389 quoted in the delivery log line, like this:
1391 => >A.N.Other <ano@some.domain> <original@ddress> ...
1393 This has been changed so that it extracts the operative address. There
1394 may be more than one such address. If so, they are comma-separated, like
1397 => >ano@some.domain,ona@other.domain <original@ddress> ...
1399 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1400 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1401 not being looked up in the rDNS to get a host name, Exim was showing the
1402 IP address twice in Received: lines, even though the IP addresses were
1403 identical. For example:
1405 Received: from [1.2.3.4] (helo=[1.2.3.4])
1407 However, if the real host name was known, it was omitting the HELO data
1408 if it matched the actual IP address. This has been tidied up so that it
1409 doesn't show the same IP address twice.
1411 PH/07 When both +timestamp and +memory debugging was on, the value given by
1412 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1413 called by the debug printing, thereby overwriting the timestamp buffer.
1414 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1416 PH/08 When the original message was included in an autoreply transport, it
1417 always said "this is a copy of the message, including all the headers",
1418 even if body_only or headers_only was set. It now gives an appropriate
1421 PH/09 Applied a patch from the Sieve maintainer which:
1423 o fixes some comments
1424 o adds the (disabled) notify extension core
1425 o adds some debug output for the result of if/elsif tests
1426 o points to the current vacation draft in the documentation
1427 and documents the missing references header update
1431 o fixes a bug in processing the envelope test (when testing
1432 multiple envelope elements, the last element determinted the
1435 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1436 Electronic Mail") by including:
1438 Auto-submitted: auto-generated
1440 in the messages that it generates (bounce messages and others, such as
1441 warnings). In the case of bounce messages for non-SMTP mesages, there was
1442 also a typo: it was using "Auto_submitted" (underscore instead of
1443 hyphen). Since every message generated by Exim is necessarily in response
1444 to another message, thes have all been changed to:
1446 Auto-Submitted: auto-replied
1448 in accordance with these statements in the RFC:
1450 The auto-replied keyword:
1452 - SHOULD be used on messages sent in direct response to another
1453 message by an automatic process,
1455 - MUST NOT be used on manually-generated messages,
1457 - MAY be used on Delivery Status Notifications (DSNs) and Message
1458 Disposition Notifications (MDNs),
1460 - MUST NOT be used on messages generated by automatic or periodic
1461 processes, except for messages which are automatic responses to
1464 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1465 to the default Received: header definition.
1467 PH/12 Added log selector acl_warn_skipped (default on).
1469 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1470 variables because (a) they are in the wrong storage pool and (b) even if
1471 they were copied, it wouldn't work properly because of the caching.
1473 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1474 checking when decoding. Apparently there are clients that generate
1475 overlong encoded strings. Why am I not surprised?
1477 PH/15 If the first argument of "${if match_address" was not empty, but did not
1478 contain an "@" character, Exim crashed. Now it writes a panic log message
1479 and treats the condition as false.
1481 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1483 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1484 extension "envelope-auth". The code is finished and in agreement with
1485 other implementations, but there is no documentation so far and in fact,
1486 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1487 not changing the active code.
1489 Print executed "if" and "elsif" statements when debugging is used. This
1490 helps a great deal to understand what a filter does.
1492 Document more things not specified clearly in RFC3028. I had all this
1493 sorted out, when out of a sudden new issues came to my mind. Oops."
1495 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1498 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1499 canonical form (as documented). However, after a host name lookup from
1500 the IP address, check_host() was doing a simple string comparison with
1501 addresses acquired from the DNS when checking that the found name did
1502 have the original IP as one of its addresses. Since any found IPv6
1503 addresses are likely to be in abbreviated form, the comparison could
1504 fail. Luckily, there already exists a function for doing the comparison
1505 by converting both addresses to binary, so now that is used instead of
1506 the text comparison.
1508 PH/20 There was another similar case to PH/19, when a complete host name was
1509 given in a host list; looking up its IP address could give an abbreviated
1510 form, whereas the current host's name might or might not be abbreviated.
1511 The same fix has been applied.
1517 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1518 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1521 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1522 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1524 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1526 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1527 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1528 Also, for "Auto-Submitted", treat anything other than "no" as
1529 non-personal, in accordance with RFC 3834. (Previously it treated
1530 anything starting "auto-" as non-personal.)
1532 TF/01 The control=submission/name=... option had a problem with syntax
1533 errors if the name included a slash character. The /name= option
1534 now slurps the rest of the string, so it can include any characters
1535 but it must come last in the list of options (after /sender_retain
1538 PH/05 Some modifications to the interface to the fake nameserver for the new
1546 TK/01 Added the "success_on_redirect" address verification option. See
1547 NewStuff for rationale and an example.
1549 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1551 PH/02 Patch to exigrep to allow it to work on syslog lines.
1553 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1554 fread() to skip over the body file's header line, because in Cygwin the
1555 header line is locked and is inaccessible.
1557 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1558 co-exist for some time) to make it clear that it is the Exim ID that is
1559 referenced, not the Message-ID: header line.
1561 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1562 string_format() function, because snprintf() does not exist on all
1565 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1566 ${stat: expansion item.
1568 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1569 protocol synchronization error", to keep the pedants happy.
1571 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1572 well as for IRIX systems, when gcc is being used. See the host.c source
1575 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1577 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1580 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1581 transport and to the smtp transport in LMTP mode.
1583 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1585 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1586 run a filter in a subprocess. This could lead to confusion in subsequent
1587 lookups in the parent process. There should also be a search_tidyup() at
1588 the end of the subprocess.
1590 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1591 only if the host matched helo_try_verify_hosts, which caused the
1592 verification to occur when the EHLO/HELO command was issued. The ACL just
1593 tested the remembered result. Now, if a previous verification attempt has
1594 not happened, "verify = helo" does it there and then.
1596 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1598 TK/03 Fix log output including CR from clamd.
1600 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1601 exist provoked a memory error which could cause a segfault.
1603 PH/15 Installed PCRE 6.2
1605 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1607 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1608 of the problem. Specifically, suggested +O2 rather than +O1 for the
1611 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1613 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1614 record, the port was not correctly logged when the outgoing_port log
1615 selector was set (it logged the transort's default port).
1617 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1618 fallback_hosts, and "hosts" in the smtp transport.
1620 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1621 host errors such as "Connection refused".
1623 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1624 authentication with radiusclient 0.4.9:
1626 - Error returned from rc_read_config was caught wrongly
1627 - Username/password not passed on to radius server due to wrong length.
1629 The presumption is that some radiusclient API changes for 4.51/PH/17
1630 were not taken care of correctly. The code is still untested by me (my
1631 Linux distribution still has 0.3.2 of radiusclient), but it was
1632 contributed by a Radius user.
1634 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1635 expanding the "port" option of the smtp transport.
1637 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1638 while reading a MIME header. Thanks to Tom Hughes for a patch.
1640 PH/24 Include config.h inside local_scan.h so that configuration settings are
1643 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1644 that in an ACL for RCPT (for example), you can examine exactly what was
1647 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1648 commands, but it was not correctly comparing the address with the actual
1649 client host address. Thus, it would show the EHLO address in Received:
1650 header lines when this was not necessary.
1652 PH/27 Added the % operator to ${eval:}.
1654 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1655 it should be ignoring failures (because with -C, for example, it has lost
1656 privilege). It wasn't ignoring creation failures other than "already
1659 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1662 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1663 because an input line was too long, either on its own, or by virtue of
1664 too many continuations, the temporary file was not being removed, and the
1665 return code was incorrect.
1667 PH/31 Missing "BOOL" in function definition in filtertest.c.
1669 PH/32 Applied Sieve patches from the maintainer.
1671 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1673 PH/33 Added "verify = not_blind".
1675 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1676 Local/Makefile (with some defaults set). These are used in built scripts
1677 such as exicyclog, but they have never been used in the exim_install
1678 script (though there are many overriding facilities there). I have
1679 arranged that the exim_install script now takes note of these two
1682 PH/35 Installed configuration files for Dragonfly.
1684 PH/36 When a locally submitted message by a trusted user did not contain a
1685 From: header, and the sender address was obtained from -f or from an SMTP
1686 MAIL command, and the trusted user did not use -F to supply a sender
1687 name, $originator_name was incorrectly used when constructing a From:
1688 header. Furthermore, $originator_name was used for submission mode
1689 messages from external hosts without From: headers in a similar way,
1690 which is clearly wrong.
1692 PH/37 Added control=suppress_local_fixups.
1694 PH/38 When log_selector = +received_sender was set, and the addition of the
1695 sender made the log line's construction buffer exactly full, or one byte
1696 less than full, an overflow happened when the terminating "\n" was
1699 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1700 when the result of a list match is failure because a DNS lookup failed.
1702 PH/40 RM_COMMAND is now used in the building process.
1704 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1705 the "build-* directories that it finds.
1707 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1708 address in a domain literal was a prefix of an interface address.
1710 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1711 when verifying a sender address, unless rewrite_headers is false.
1713 PH/44 Wrote a long comment about why errors_to addresses are verified as
1714 recipients, not senders.
1716 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1717 the ratelimit ACL was added.
1719 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1721 PH/46 Added extra information about PostgreSQL errors to the error string.
1723 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1724 avoiding the need to install special zones in a real server. This is
1725 backwards compatible; if it can't find the fake resolver, it drops back.
1726 Thus, both old and new test suites can be run.
1728 TF/02 Added util/ratelimit.pl
1730 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1731 clock is set back in time.
1733 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1734 Candler <B.Candler@pobox.com>.
1736 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1737 OK for addresses that are the result of redirections.
1739 PH/48 A number of further additions for the benefit of the new test suite,
1740 including a fake gethostbyname() that interfaces to the fake DNS resolver
1743 TF/06 The fix for widen_domains has also been applied to qualify_single and
1744 search_parents which are the other dnslookup options that can cause
1747 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1750 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1752 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1753 service" if it encountered an MX record with an empty target hostname.
1754 The message is now "an MX or SRV record indicated no SMTP service".
1756 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1757 if the DNS of the sending site is misconfigured. This is quite a
1758 common situation. This change restores the behaviour of treating a
1759 helo verification defer as a failure.
1761 PH/51 If self=fail was set on a router, the bounce message did not include the
1762 actual error message.
1768 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1770 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1771 command itself ended in error, the underlying message about the transport
1772 filter timeout was being overwritten with the pipe command error. Now the
1773 underlying error message should be appended to the second error message.
1775 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1777 PH/02 Reduce the amount of output that "make" produces by default. Full output
1778 can still be requested.
1780 PH/03 The warning log line about a condition test deferring for a "warn" verb
1781 was being output only once per connection, rather than after each
1782 occurrence (because it was using the same function as for successful
1783 "warn" verbs). This seems wrong, so I have changed it.
1785 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1786 it should not have, which might have caused a crash in the right
1787 circumstances, but probably never did.
1789 PH/04 Installed a modified version of Tony Finch's patch to make submission
1790 mode fix the return path as well as the Sender: header line, and to
1791 add a /name= option so that you can make the user's friendly name appear
1794 TF/03 Added the control = fakedefer ACL modifier.
1796 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1797 Mark Lowes for thorough testing.
1799 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1801 TK/03 Merged latest SRS patch from Miles Wilton.
1803 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1804 with the definition in sysexits.h (which is #included earlier).
1805 Fortunately, Exim does not actually use EX_OK. The code used to try to
1806 preserve the sysexits.h value, by assumimg that macro definitions were
1807 scanned for macro replacements. I have been disabused of this notion,
1808 so now the code just undefines EX_OK before #including unistd.h.
1810 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1811 in the smtp transport. When a block could not be written in a single
1812 write() function, the timeout was being re-applied to each part-write.
1813 This seems wrong - if the receiver was accepting one byte at a time it
1814 would take for ever. The timeout is now adjusted when this happens. It
1815 doesn't have to be particularly precise.
1817 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1818 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1820 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1821 without a domain if the check to <postmaster@domain> fails.
1823 SC/01 Eximstats: added -xls and the ability to specify output files
1824 (patch written by Frank Heydlauf).
1826 SC/02 Eximstats: use FileHandles for outputing results.
1828 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1830 SC/04 Eximstats: fixed display of large numbers with -nvr option
1832 SC/05 Eximstats: fixed merging of reports with empty tables.
1834 SC/06 Eximstats: added the -include_original_destination flag
1836 SC/07 Eximstats: removed tabs and trailing whitespace.
1838 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1840 TK/06 MBOX spool code: Add real "From " MBOX separator line
1841 so the .eml file is really in mbox format (even though
1842 most programs do not really care). Patch from Alex Miller.
1844 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1845 The latter is generated from $received_to and is only set if the
1846 message has one envelope recipient. SA can use these headers,
1847 obviously out-of-the-box. Patch from Alex Miller.
1849 PH/08 The ${def test on a variable was returning false if the variable's
1850 value was "0", contrary to what the specification has always said!
1851 The result should be true unless the variable is empty.
1853 PH/09 The syntax error of a character other than { following "${if
1854 def:variable_name" (after optional whitespace) was not being diagnosed.
1855 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1856 accidental colon was present, for example, could give incorrect results.
1858 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1859 result is used (not including appendfile, where other changes are about
1862 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1863 This involved changing a lot of size variables from int to off_t. It
1864 should work with maildirs and everything.
1866 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1867 spamd dying while we are connected to it.
1869 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1872 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1873 function for BDB 4.3.
1875 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1877 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1878 the sources for PCRE have been re-arranged and more files are now
1881 PH/15 The code I had for printing potentially long long variables in PH/11
1882 above was not the best (it lost precision). The length of off_t variables
1883 is now inspected at build time, and an appropriate printing format (%ld
1884 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1885 to be "long long int" or "long int". This is needed for the internal
1886 formatting function string_vformat().
1888 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1889 the configuration file to be ":syslog", then the script "guesses" where
1890 the logs files are, rather than using the compiled in default. In our
1891 case the guess is not the same as the compiled default, so the script
1892 suddenly stopped working when I started to use syslog. The patch checks
1893 to see if log_file_path is "". If so, it attempts to read it from exim
1894 with no configuration file to get the compiled in version, before it
1895 falls back to the previous guessing code."
1897 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1898 implementing BATV in an Exim configuration. See NewStuff for the gory
1901 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1902 Makefile that are specific to HP-UX.
1904 PH/18 If the "use_postmaster" option was set for a recipient callout together
1905 with the "random" option, the postmaster address was used as the MAIL
1906 FROM address for the random test, but not for the subsequent recipient
1907 test. It is now used for both.
1909 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1910 patch removes a few documentation additions to RFC 3028, because the
1911 latest draft now contains them. It adds the new en;ascii-case comparator
1912 and a new error check for 8bit text in MIME parts. Comparator and
1913 require names are now matched exactly. I enabled the subaddress
1914 extension, but it is not well tested yet (read: it works for me)."
1916 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1917 rework some of the code of TK/09 above to avoid the hardwired use of
1918 "%lld" and "long long". Replaced the call to snprintf() with a call to
1921 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1922 records point to non-existent hosts", "retry timeout exceeded", and
1923 "retry time not reached for any host after a long failure period".
1925 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1926 experimental DomainKeys support:
1928 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1929 (2) On an error such as an illegally used "control", the wrong name for
1930 the control was given.
1932 These problems did NOT occur unless DomainKeys support was compiled.
1934 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1936 PH/24 Added ${if match_ip condition.
1938 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1939 databases so that it will be absolutely obvious if a crash occurs in the
1940 DB library. This is a regular occurrence (often caused by mis-matched
1943 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1944 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1945 picked up on a user's system that detects such things. There doesn't seem
1946 to be a gcc warning option for this - only an attribute that has to be
1947 put on the function's prototype. It seems that in Fedora Core 4 they have
1948 set this on a number of new functions. No doubt there will be more in due
1951 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1952 specify a transport. However, if an address that was verified by such a
1953 router was the subject of a callout, Exim crashed because it tried to
1954 read the rcpt_include_affixes from the non-existent transport. Now it
1955 just assumes that the setting of that option is false. This bug was
1956 introduced by 4.51/PH/31.
1958 PH/28 Changed -d+all to exclude +memory, because that information is very
1959 rarely of interest, but it makes the output a lot bigger. People tend to
1960 do -d+all out of habit.
1962 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1963 code in os-type was giving problems when libc.so lives in lib64, like on
1966 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1967 aren't the modern standard, and it seems that some systems' include files
1968 don't always have them. Exim was already checking for some of the newer
1969 ones like T_AAAA, and defining it itself. I've added checks for all the
1970 record types that Exim uses.
1972 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1973 not automatically generating a new one, as it is supposed to. This
1974 prevented TLS from working. If the file did exist, but contained invalid
1975 data, a new version was generated, as expected. It was only the case of a
1976 non-existent file that was broken.
1978 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1979 with a change in libdomainkeys > 0.64.
1981 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1982 from DNS. If the selector record carries the flag, it now has
1983 precedence over the domain-wide flag.
1985 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1987 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1988 the use of a port name that isn't defined in /etc/services) occurred, the
1989 message was deferred as in a normal delivery, and thus remained on the
1990 spool, instead of being failed because of the mua_wrapper setting. This
1991 is now fixed, and I tidied up some of the mua_wrapper messages at the
1994 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1995 the messages in a hash of arrays rather than using individual hashes.
1996 This is a bit cleaner and results in dramatic memory savings, albeit
1997 at a slight CPU cost.
1999 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2000 as requested by Marc Sherman.
2002 SC/10 Eximstats: added histograms for user specified patterns as requested
2005 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2007 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2008 fopen() in the content-scanning modules that did not already have it.
2014 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2015 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2017 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2019 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2021 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2022 to test Sieve filters that use "vacation".
2024 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2025 that changes the way the GnuTLS parameters are stored in the cache file.
2026 The new format can be generated externally. For backward compatibility,
2027 if the data in the cache doesn't make sense, Exim assumes it has read an
2028 old-format file, and it generates new data and writes a new file. This
2029 means that you can't go back to an older release without removing the
2032 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2033 work if there are any delivery delays because "one_time" forces the
2034 parent to be marked "delivered", so its unseen clone is never tried
2035 again. For this reason, Exim now forbids the simultaneous setting of
2038 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2039 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2040 entry if you want to know the details. The fix, however, neglected to
2041 consider the case when local delivery batching is involved. The test for
2042 "previously delivered" was not happening when checking to see if an
2043 address could be batched with a previous (undelivered) one; under
2044 certain circumstances this could lead to multiple deliveries to the same
2047 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2048 in its include files, and this causes problems building Exim.
2050 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2051 header_syntax) but Exim was just ignoring anything given after a slash.
2052 In particular, this caused confusion with an attempt to use "verify =
2053 reverse_host_lookup/defer_ok". An error is now given when options are
2054 supplied for verify items that do not have them. (Maybe reverse_host_
2055 lookup should have a defer_ok option, but that's a different point.)
2057 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2058 defined by RFC 821) to 2048, because there were problems with some AUTH
2059 commands, and RFC 1869 says the size should be increased for extended
2060 SMTP commands that take arguments.
2062 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2065 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2066 "unknown" error; now it says that the functionality isn't in the binary.
2068 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2069 an address' error message when a string expansion fails (syntax or
2070 whatever). Otherwise the password may appear in the log. Following change
2071 PH/42 below, there is no longer a chance of it appearing in a bounce
2074 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2076 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2077 of its addresses was ever tried. (Bugzilla bug #2.)
2079 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2080 the result incorrectly in the debug output. (It correctly added a newline
2081 to what was transported.)
2083 TF/01 Added $received_time.
2085 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2086 commented out examples of how to interface to a virus scanner and to
2087 SpamAssassin. Also added commented examples of av_scanner and
2088 spamd_address settings.
2090 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2091 and controls are allowed in which ACLs. There were a couple of minor
2092 errors. Some of the entries in the conditions table (which is a table of
2093 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2094 negation of where the condition IS allowed.
2096 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2098 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2099 header file does not have a version number, so I've had to invent a new
2100 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2101 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2102 radiusclient), but it was contributed by a Radius user.
2104 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2105 files or directories, for OpenSSL.
2107 PH/19 When an Exim process that is running as root has to create an Exim log
2108 file, it does so in a subprocess that runs as exim:exim so as to get the
2109 ownership right at creation (otherwise, other Exim processes might see
2110 the file with the wrong ownership). There was no test for failure of this
2111 fork() call, which would lead to the process getting stuck as it waited
2112 for a non-existent subprocess. Forks do occasionally fail when resources
2113 run out. I reviewed all the other calls to fork(); they all seem to check
2116 PH/20 When checking for unexpected SMTP input at connect time (before writing
2117 the banner), Exim was not dealing correctly with a non-positive return
2118 from the read() function. If the client had disconnected by this time,
2119 the result was a log entry for a synchronization error with an empty
2120 string after "input=" when read() returned zero. If read() returned -1
2121 (an event I could not check), uninitialized data bytes were printed.
2122 There were reports of junk text (parts of files, etc) appearing after
2125 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2127 PH/22 Added support for macro redefinition, and (re)definition in between
2128 driver and ACL definitions.
2130 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2131 forgetting to use the resulting value; it was using the unexpanded value.
2133 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2134 hadn't been configured. The fix is from Juergen Kreileder, who
2135 understands it better than I do:
2137 "Here's what I see happening with three configured cyrus_sasl
2138 authenticators configured (plain, login, cram-md5):
2140 On startup auth_cyrus_sasl_init() gets called for each of these.
2141 This means three calls to sasl_listmech() without a specified mech_list.
2142 => SASL tests which mechs of all available mechs actually work
2143 => three warnings about OTP not working
2144 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2146 With the patch, sasl_listmech() also gets called three times. But now
2147 SASL's mech_list option is set to the server_mech specified in the the
2148 authenticator. Or in other words, the answer from sasl_listmech()
2149 gets limited to just the mech you're testing for (which is different
2151 => the return list contains just 'plain' or 'login', 'cram-md5' or
2152 nothing depending on the value of ob->server_mech.
2154 I've just tested the patch: Authentication still works fine,
2155 unavailable mechs specified in the exim configuration are still
2156 caught, and the auth.log warnings about OTP are gone."
2158 PH/25 When debugging is enabled, the contents of the command line are added
2159 to the debugging output, even when log_selector=+arguments is not
2162 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2163 answer is "GNU", and only if the return is "GNU/something" is the answer
2166 PH/27 $acl_verify_message is now set immediately after the failure of a
2167 verification in an ACL, and so is available in subsequent modifiers. In
2168 particular, the message can be preserved by coding like this:
2170 warn !verify = sender
2171 set acl_m0 = $acl_verify_message
2173 Previously, $acl_verify_message was set only while expanding "message"
2174 and "log_message" when a very denied access.
2176 PH/28 Modified OS/os.c-Linux with
2178 -#ifndef OS_LOAD_AVERAGE
2179 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2181 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2182 nomenclature these days.)
2184 PH/29 Installed patch from the Sieve maintainer that adds the options
2185 sieve_useraddress and sieve_subaddress to the redirect router.
2187 PH/30 In these circumstances:
2188 . Two addresses routed to the same list of hosts;
2189 . First host does not offer TLS;
2190 . First host accepts first address;
2191 . First host gives temporary error to second address;
2192 . Second host offers TLS and a TLS session is established;
2193 . Second host accepts second address.
2194 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2195 and peerdn, if requested) that were in fact used only for the second
2198 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2199 attention to any local part prefix or suffix that was matched by the
2200 router that accepted the address. It now behaves in the same way as it
2201 does for delivery: the affixes are removed from the local part unless
2202 rcpt_include_affixes is set on the transport.
2204 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2205 timeout during the DATA phase of an incoming message.
2207 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2208 applied a patch sanctioned by the Sieve maintainer.
2210 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2211 the uid or gid is negative. A case of a negative gid caused this to be
2212 noticed. The fix allows for either to be negative.
2214 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2215 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2218 PH/36 The change PH/12 above was broken. Fixed it.
2220 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2221 the grounds that routing the same address twice would always produce the
2222 same answer. This might have been true once, but it is certainly no
2223 longer true now. Routing a child address may depend on the previous
2224 routing that produced that child. Some complicated redirection strategies
2225 went wrong when messages had multiple recipients, and made Exim's
2226 behaviour dependent on the order in which the addresses were given.
2228 I have moved the duplicate checking until after the routing is complete.
2229 Exim scans the addresses that are assigned to local and remote
2230 transports, and removes any duplicates. This means that more work will be
2231 done, as duplicates will always all be routed, but duplicates are
2232 presumably rare, so I don't expect this is of any significance.
2234 For deliveries to pipes, files, and autoreplies, the duplicate checking
2235 still happens during the routing process, since they are not going to be
2238 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2239 It corrects a timeout issue with spamd. This is Ian's comment: "The
2240 background is that sometimes spamd either never reads data from a
2241 connection it has accepted, or it never writes response data. The exiscan
2242 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2243 blindly assumes that writes won't block so it may never time out."
2245 PH/39 Allow G after quota size as well as K and M.
2247 PH/40 The value set for $authenticated_id in an authenticator may not contain
2248 binary zeroes or newlines because the value is written to log lines and
2249 to spool files. There was no check on this. Now the value is run through
2250 the string_printing() function so that such characters are converted to
2251 printable escape sequences.
2253 PH/41 $message_linecount is a new variable that contains the total number of
2254 lines in the message. Compare $body_linecount, which is the count for the
2257 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2258 bounce and delay warning messages, except in certain special cases, which
2261 (a) An SMTP error message from a remote host;
2262 (b) A message specified in a :fail: redirection;
2263 (c) A message specified in a "fail" command in a system filter;
2264 (d) A message specified in a FAIL return from the queryprogram router;
2265 (e) A message specified by the cannot_route_message router option.
2267 In these cases only, Exim does include the error details in bounce and
2268 warning messages. There are also a few cases where bland messages such
2269 as "unrouteable address" or "local delivery error" are given.
2271 PH/43 $value is now also set for the "else" part of a ${run expansion.
2273 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2274 being worked on, but at least Exim now implements the latest version to
2277 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2278 process to complete was treated as a delivery failure, a timeout while
2279 writing the message to the pipe was logged, but erroneously treated as a
2280 successful delivery. Such timeouts include transport filter timeouts. For
2281 consistency with the overall process timeout, these timeouts are now
2282 treated as errors, giving rise to delivery failures by default. However,
2283 there is now a new Boolean option for the pipe transport called
2284 timeout_defer, which, if set TRUE, converts the failures into defers for
2285 both kinds of timeout. A transport filter timeout is now identified in
2288 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2289 systems where "make" and "gmake" are different, calling "gmake" at top
2290 level broke things. I've arranged for the value of $(MAKE) to be passed
2291 from the Makefile to this script so that it can call the same version of
2295 A note about Exim versions 4.44 and 4.50
2296 ----------------------------------------
2298 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2299 changes of various kinds. As a consequence, a big documentation update was
2300 needed. This delayed the release for rather longer than seemed good, especially
2301 in the light of a couple of (minor) security issues. Therefore, the changes
2302 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2303 release. So 4.44 and 4.50 are in effect two different branches that both start
2306 I have left the 4.50 change log unchanged; it contains all the changes since
2307 4.43. The change log for 4.44 is below; many of its items are identical to
2308 those for 4.50. This seems to be the most sensible way to preserve the
2309 historical information.
2315 1. Minor wording change to the doc/README.SIEVE file.
2317 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2318 computation of the current number of files was incorrect.
2320 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2321 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2324 4. Give more explanation in the error message when the command for a transport
2325 filter fails to execute.
2327 5. There are several places where Exim runs a non-Exim command in a
2328 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2329 was being done only for the command run by the queryprogram router. It is
2330 now done for all such subprocesses. The other cases are: ${run, transport
2331 filters, and the commands run by the lmtp and pipe transports.
2333 6. Added CONFIGURE_GROUP build-time option.
2335 7. Some older OS have a limit of 256 on the maximum number of file
2336 descriptors. Exim was using setrlimit() to set 1000 as a large value
2337 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2338 systems. I've change it so that if it can't get 1000, it tries for 256.
2340 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2341 was an oversight, and furthermore, ever since the addition of extra
2342 controls (e.g. 4.43/32), the checks on when to allow different forms of
2343 "control" were broken. There should now be diagnostics for all cases when a
2344 control that does not make sense is encountered.
2346 9. Added the /retain_sender option to "control=submission".
2348 10. $recipients is now available in the predata ACL (oversight).
2350 11. Tidy the search cache before the fork to do a delivery from a message
2351 received from the command line. Otherwise the child will trigger a lookup
2352 failure and thereby defer the delivery if it tries to use (for example) a
2353 cached ldap connection that the parent has called unbind on.
2355 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2356 of $address_data from the recipient verification was clobbered by the
2357 sender verification.
2359 13. The value of address_data from a sender verification is now available in
2360 $sender_address_data in subsequent conditions in the ACL statement.
2362 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2364 15. Added a new option "connect=<time>" to callout options, to set a different
2367 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2368 was its contents. (It was OK if the option was not defined at all.)
2370 17. A "Completed" log line is now written for messages that are removed from
2371 the spool by the -Mrm option.
2373 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2374 information about exactly what failed.
2376 19. Added -dd to debug only the daemon process.
2378 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2379 handles timeouts, both on the server side and network timeouts. Renamed the
2380 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2382 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2383 It is now set to "smtps".
2385 22. $host_address is now set to the target address during the checking of
2386 ignore_target_hosts.
2388 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2389 being passed; this would have caused $sender_host_name to have been used if
2390 matching the list had actually called for a host name (not very likely,
2391 since this list is usually IP addresses). A host name is now passed as
2394 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2395 code that specifies a non-restarting handler (typically sigaction() in
2396 modern systems) in an attempt to fix a rare and obscure crash bug.
2398 25. Narrowed the window for a race in the daemon that could cause it to ignore
2399 SIGCHLD signals. This is not a major problem, because they are used only to
2400 wake it up if nothing else does.
2402 26. A malformed maildirsize file could cause Exim to calculate negative values
2403 for the mailbox size or file count. Odd effects could occur as a result.
2404 The maildirsize information is now recalculated if the size or filecount
2407 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2408 support for a long time. Removed HAVE_SYS_VFS_H.
2410 28. Installed the latest version of exipick from John Jetmore.
2412 29. In an address list, if the pattern was not a regular expression, an empty
2413 subject address (from a bounce message) matched only if the pattern was an
2414 empty string. Non-empty patterns were not even tested. This was the wrong
2415 because it is perfectly reasonable to use an empty address as part of a
2416 database query. An empty address is now tested by patterns that are
2417 lookups. However, all the other forms of pattern expect the subject to
2418 contain a local part and a domain, and therefore, for them, an empty
2419 address still always fails if the pattern is not itself empty.
2421 30. Exim went into a mad DNS loop when attempting to do a callout where the
2422 host was specified on an smtp transport, and looking it up yielded more
2423 than one IP address.
2425 31. Re-factored the code for checking spool and log partition space into a
2426 function that finds that data and another that does the check. The former
2427 is then used to implement four new variables: $spool_space, $log_space,
2428 $spool_inodes, and $log_inodes.
2430 32. The RFC2047 encoding function was originally intended for short strings
2431 such as real names; it was not keeping to the 75-character limit for
2432 encoded words that the RFC imposes. It now respects the limit, and
2433 generates multiple encoded words if necessary. To be on the safe side, I
2434 have increased the buffer size for the ${rfc2047: expansion operator from
2437 33. It is now permitted to omit both strings after an "if" condition; if the
2438 condition is true, the result is "true". As before, when the second string
2439 is omitted, a false condition yields an empty string. This makes it less
2440 cumbersome to write custom ACL and router conditions.
2442 34. Failure to deliver a bounce message always caused it to be frozen, even if
2443 there was an errors_to setting on the router. The errors_to setting is now
2446 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2447 canonical form (fully expanded) before being placed in
2448 $sender_host_address.
2450 36. The table in the code that translates DNS record types into text (T_A to
2451 "A" for instance) was missing entries for NS and CNAME. It is just possible
2452 that this could have caused confusion if both these types were looked up
2453 for the same domain, because the text type is used as part of Exim's
2454 per-process caching. But the chance of anyone hitting this buglet seems
2457 37. The dnsdb lookup has been extended in a number of ways.
2459 (1) There is a new type, "zns", which walks up the domain tree until it
2460 finds some nameserver records. It should be used with care.
2462 (2) There is a new type, "mxh", which is like "mx" except that it returns
2463 just the host names, not the priorities.
2465 (3) It is now possible to give a list of domains (or IP addresses) to be
2466 looked up. The behaviour when one of the lookups defers can be
2467 controlled by a keyword.
2469 (4) It is now possible to specify the separator character for use when
2470 multiple records are returned.
2472 38. The dnslists ACL condition has been extended: it is now possible to supply
2473 a list of IP addresses and/or domains to be looked up in a particular DNS
2476 39. Added log_selector=+queue_time_overall.
2478 40. When running the queue in the test harness, wait just a tad after forking a
2479 delivery process, to get repeatability of debugging output.
2481 41. Include certificate and key file names in error message when GnuTLS fails
2482 to set them up, because the GnuTLS error message doesn't include the name
2483 of the failing file when there is a problem reading it.
2485 42. Allow both -bf and -bF in the same test run.
2487 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2489 44. The "Exiscan patch" is now merged into the mainline Exim source.
2491 45. Sometimes the final signoff response after QUIT could fail to get
2492 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2493 < 0 before doing a fflush(). This bug looks as though it goes back to the
2494 introduction of TLS in release 3.20, but "sometimes" must have been rare
2495 because the tests only now provoked it.
2497 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2498 (this can affect the format of dates).
2500 47. exim_tidydb, when checking for the continued existence of a message for
2501 which it has found a message-specific retry record, was not finding
2502 messages that were in split spool directories. Consequently, it was
2503 deleting retry records that should have stayed in existence.
2505 48. Steve fixed some bugs in eximstats.
2507 49. The SPA authentication driver was not abandoning authentication and moving
2508 on to the next authenticator when an expansion was forced to fail,
2509 contradicting the general specification for all authenticators. Instead it
2510 was generating a temporary error. It now behaves as specified.
2512 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2513 (the order specifies the preference for clients). The order is now AES256,
2514 AES128, 3DES, ARCFOUR128.
2516 51. Small patch to Sieve code - explicitly set From: when generating an
2519 52. Exim crashed if a remote delivery caused a very long error message to be
2520 recorded - for instance if somebody sent an entire SpamAssassin report back
2521 as a large number of 550 error lines. This bug was coincidentally fixed by
2522 increasing the size of one of Exim's internal buffers (big_buffer) that
2523 happened as part of the Exiscan merge. However, to be on the safe side, I
2524 have made the code more robust (and fixed the comments that describe what
2527 53. Now that there can be additional text after "Completed" in log lines (if
2528 the queue_time_overall log selector is set), a one-byte patch to exigrep
2529 was needed to allow it to recognize "Completed" as not the last thing in
2532 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2533 patch that reportedly fixes this has been added. I am not expert enough to
2534 create a test for it. This is what the patch creator wrote:
2536 "I found a little strange behaviour of ldap code when working with
2537 Windows 2003 AD Domain, where users was placed in more than one
2538 Organization Units. When I tried to give exim partial DN, the exit code
2539 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2540 But simultaneously result of request was absolutely normal ldap result,
2541 so I produce this patch..."
2543 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2544 REFERENCE, so I have modified the code to exclude the patch when that macro
2547 55. Some experimental protocols are using DNS PTR records for new purposes. The
2548 keys for these records are domain names, not reversed IP addresses. The
2549 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2550 leaves it alone. Component reversal etc. now happens only for IP addresses.
2553 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2555 57. Double the size of the debug message buffer (to 2048) so that more of very
2556 long debug lines gets shown.
2558 58. The exicyclog utility now does better if the number of log files to keep
2559 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2561 59. Two changes related to the smtp_active_hostname option:
2563 (1) $smtp_active_hostname is now available as a variable.
2564 (2) The default for smtp_banner uses $smtp_active_hostname instead
2565 of $primary_hostname.
2567 60. The host_aton() function is supposed to be passed a string that is known
2568 to be a valid IP address. However, in the case of IPv6 addresses, it was
2569 not checking this. This is a hostage to fortune. Exim now panics and dies
2570 if the condition is not met. A case was found where this could be provoked
2571 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2572 components; fortuitously, this particular loophole had already been fixed
2573 by change 4.50/55 above.
2575 If there are any other similar loopholes, the new check in host_aton()
2576 itself should stop them being exploited. The report I received stated that
2577 data on the command line could provoke the exploit when Exim was running as
2578 exim, but did not say which command line option was involved. All I could
2579 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2580 running as the user.
2583 61. There was a buffer overflow vulnerability in the SPA authentication code
2584 (which came originally from the Samba project). I have added a test to the
2585 spa_base64_to_bits() function which I hope fixes it.
2588 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2589 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2591 63. The daemon start-up calls getloadavg() while still root for those OS that
2592 need the first call to be done as root, but it missed one case: when
2593 deliver_queue_load_max is set with deliver_drop_privilege. This is
2594 necessary for the benefit of the queue runner, because there is no re-exec
2595 when deliver_drop_privilege is set.
2597 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2598 This has been fixed.
2600 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2601 was in use, was not putting the data itself into the right store pool;
2602 consequently, it could be overwritten for a subsequent message in the same
2603 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2606 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2608 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2609 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2610 the function were treating the return as a boolean value, which happened to
2611 work because 0=false and not-0=true, but is not correct code.
2613 68. The host_aton() function was not handling scoped IPv6 addresses (those
2614 with, for example, "%eth0" on the end) correctly.
2616 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2617 negated items (that is, ~something) in unsigned ints. Some compilers
2618 apparently mutter when there is no cast.
2620 70. If an address verification called from an ACL failed, and did not produce a
2621 user-specific message (i.e. there was only a "system" message), nothing was
2622 put in $acl_verify_message. In this situation, it now puts the system
2625 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2626 session; change 4.31/43 added the unwanted input to the log line - except
2627 that it did not do this in the start of session case. It now does.
2629 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2630 This is wrong and can cause the other end to generate a synchronization
2631 error if it is another Exim or anything else that does the synchronization
2632 check. A QUIT command is no longer sent after a timeout.
2634 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2635 during host lookups.
2637 74. The defer_ok option of callout verification was not working if it was used
2638 when verifying addresses in header lines, that is, for this case:
2640 verify = header_sender/callout=defer_ok
2642 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2643 those file descriptors could be used for SMTP connections. If anything
2644 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2645 could be sent to the SMTP client, causing chaos. The daemon now opens
2646 stdin, stdout, and stderr to /dev/null when it puts itself into the
2649 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2650 log by default. The user can override this with suitable Perl magic.
2652 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2653 the log message when discard triggers, was not working for the DATA ACL or
2654 for the non-SMTP ACL.
2656 78. Error message wording change in sieve.c.
2658 79. If smtp_accept_max_per_host was set, the number of connections could be
2659 restricted to fewer than expected, because the daemon was trying to set up
2660 a new connection before checking whether the processes handling previous
2661 connections had finished. The check for completed processes is now done
2662 earlier. On busy systems, this bug wouldn't be noticed because something
2663 else would have woken the daemon, and it would have reaped the completed
2666 80. If a message was submitted locally by a user whose login name contained one
2667 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2668 It caused a spool format error. I have fixed the spool reading code. A
2669 related problem was that the "from" clause in the Received: line became
2670 illegal because of the space(s). It is now covered by ${quote_local_part.
2672 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2675 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2677 83. Patch from Sieve maintainer to fix unterminated string problem in
2678 "vacation" handling.
2680 84. Some minor changes to the Linux configuration files to help with other
2681 OS variants using glibc.
2683 85. One more patch for Sieve to update vacation handling to latest spec.
2686 ----------------------------------------------------
2687 See the note above about the 4.44 and 4.50 releases.
2688 ----------------------------------------------------
2694 1. Change 4.43/35 introduced a bug that caused file counts to be
2695 incorrectly computed when quota_filecount was set in an appendfile
2698 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2699 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2702 3. Give more explanation in the error message when the command for a transport
2703 filter fails to execute.
2705 4. There are several places where Exim runs a non-Exim command in a
2706 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2707 was being done only for the command run by the queryprogram router. It is
2708 now done for all such subprocesses. The other cases are: ${run, transport
2709 filters, and the commands run by the lmtp and pipe transports.
2711 5. Some older OS have a limit of 256 on the maximum number of file
2712 descriptors. Exim was using setrlimit() to set 1000 as a large value
2713 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2714 systems. I've change it so that if it can't get 1000, it tries for 256.
2716 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2717 was an oversight, and furthermore, ever since the addition of extra
2718 controls (e.g. 4.43/32), the checks on when to allow different forms of
2719 "control" were broken. There should now be diagnostics for all cases when a
2720 control that does not make sense is encountered.
2722 7. $recipients is now available in the predata ACL (oversight).
2724 8. Tidy the search cache before the fork to do a delivery from a message
2725 received from the command line. Otherwise the child will trigger a lookup
2726 failure and thereby defer the delivery if it tries to use (for example) a
2727 cached ldap connection that the parent has called unbind on.
2729 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2730 of $address_data from the recipient verification was clobbered by the
2731 sender verification.
2733 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2734 was its contents. (It was OK if the option was not defined at all.)
2736 11. A "Completed" log line is now written for messages that are removed from
2737 the spool by the -Mrm option.
2739 12. $host_address is now set to the target address during the checking of
2740 ignore_target_hosts.
2742 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2743 being passed; this would have caused $sender_host_name to have been used if
2744 matching the list had actually called for a host name (not very likely,
2745 since this list is usually IP addresses). A host name is now passed as
2748 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2749 code that specifies a non-restarting handler (typically sigaction() in
2750 modern systems) in an attempt to fix a rare and obscure crash bug.
2752 15. Narrowed the window for a race in the daemon that could cause it to ignore
2753 SIGCHLD signals. This is not a major problem, because they are used only to
2754 wake it up if nothing else does.
2756 16. A malformed maildirsize file could cause Exim to calculate negative values
2757 for the mailbox size or file count. Odd effects could occur as a result.
2758 The maildirsize information is now recalculated if the size or filecount
2761 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2762 support for a long time. Removed HAVE_SYS_VFS_H.
2764 18. Updated exipick to current release from John Jetmore.
2766 19. Allow an empty sender to be matched against a lookup in an address list.
2767 Previously the only cases considered were a regular expression, or an
2770 20. Exim went into a mad DNS lookup loop when doing a callout where the
2771 host was specified on the transport, if the DNS lookup yielded more than
2774 21. The RFC2047 encoding function was originally intended for short strings
2775 such as real names; it was not keeping to the 75-character limit for
2776 encoded words that the RFC imposes. It now respects the limit, and
2777 generates multiple encoded words if necessary. To be on the safe side, I
2778 have increased the buffer size for the ${rfc2047: expansion operator from
2781 22. Failure to deliver a bounce message always caused it to be frozen, even if
2782 there was an errors_to setting on the router. The errors_to setting is now
2785 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2786 canonical form (fully expanded) before being placed in
2787 $sender_host_address.
2789 24. Updated eximstats to version 1.33
2791 25. Include certificate and key file names in error message when GnuTLS fails
2792 to set them up, because the GnuTLS error message doesn't include the name
2793 of the failing file when there is a problem reading it.
2795 26. Expand error message when OpenSSL has problems setting up cert/key files.
2798 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2799 (this can affect the format of dates).
2801 28. exim_tidydb, when checking for the continued existence of a message for
2802 which it has found a message-specific retry record, was not finding
2803 messages that were in split spool directories. Consequently, it was
2804 deleting retry records that should have stayed in existence.
2806 29. eximstats updated to version 1.35
2807 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2808 1.35 - bugfix such that pie charts by volume are generated correctly
2810 30. The SPA authentication driver was not abandoning authentication and moving
2811 on to the next authenticator when an expansion was forced to fail,
2812 contradicting the general specification for all authenticators. Instead it
2813 was generating a temporary error. It now behaves as specified.
2815 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2816 (the order specifies the preference for clients). The order is now AES256,
2817 AES128, 3DES, ARCFOUR128.
2819 31. Small patch to Sieve code - explicitly set From: when generating an
2822 32. Exim crashed if a remote delivery caused a very long error message to be
2823 recorded - for instance if somebody sent an entire SpamAssassin report back
2824 as a large number of 550 error lines. This bug was coincidentally fixed by
2825 increasing the size of one of Exim's internal buffers (big_buffer) that
2826 happened as part of the Exiscan merge. However, to be on the safe side, I
2827 have made the code more robust (and fixed the comments that describe what
2830 33. Some experimental protocols are using DNS PTR records for new purposes. The
2831 keys for these records are domain names, not reversed IP addresses. The
2832 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2833 leaves it alone. Component reversal etc. now happens only for IP addresses.
2836 34. The host_aton() function is supposed to be passed a string that is known
2837 to be a valid IP address. However, in the case of IPv6 addresses, it was
2838 not checking this. This is a hostage to fortune. Exim now panics and dies
2839 if the condition is not met. A case was found where this could be provoked
2840 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2841 components; fortuitously, this particular loophole had already been fixed
2842 by change 4.50/55 or 4.44/33 above.
2844 If there are any other similar loopholes, the new check in host_aton()
2845 itself should stop them being exploited. The report I received stated that
2846 data on the command line could provoke the exploit when Exim was running as
2847 exim, but did not say which command line option was involved. All I could
2848 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2849 running as the user.
2852 35. There was a buffer overflow vulnerability in the SPA authentication code
2853 (which came originally from the Samba project). I have added a test to the
2854 spa_base64_to_bits() function which I hope fixes it.
2857 36. The daemon start-up calls getloadavg() while still root for those OS that
2858 need the first call to be done as root, but it missed one case: when
2859 deliver_queue_load_max is set with deliver_drop_privilege. This is
2860 necessary for the benefit of the queue runner, because there is no re-exec
2861 when deliver_drop_privilege is set.
2863 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2864 was in use, was not putting the data itself into the right store pool;
2865 consequently, it could be overwritten for a subsequent message in the same
2866 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2869 38. Sometimes the final signoff response after QUIT could fail to get
2870 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2871 < 0 before doing a fflush(). This bug looks as though it goes back to the
2872 introduction of TLS in release 3.20, but "sometimes" must have been rare
2873 because the tests only now provoked it.
2879 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2880 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2881 Now it returns an integer. A number of calls were still expecting a T/F
2882 return. Fortuitously, in all cases, the tests worked in OK situations,
2883 which is the norm. However, things would have gone wrong on any write
2884 failures on the smtp file descriptor. This function is used when sending
2885 messages over SMTP and also when doing verify callouts.
2887 2. When Exim is called to do synchronous delivery of a locally submitted
2888 message (the -odf or -odi options), it no longer closes stderr before doing
2891 3. Implemented the mua_wrapper option.
2893 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2895 5. Implemented the functions header_remove(), header_testname(),
2896 header_add_at_position(), and receive_remove_recipient(), and exported them
2899 6. If an ACL "warn" statement specified the addition of headers, Exim already
2900 inserted X-ACL-Warn: at the start if there was no header name. However, it
2901 was not making this test for the second and subsequent header lines if
2902 there were newlines in the string. This meant that an invalid header could
2903 be inserted if Exim was badly configured.
2905 7. Allow an ACL "warn" statement to add header lines at the start or after all
2906 the Received: headers, as well as at the end.
2908 8. Added the rcpt_4xx retry error code.
2910 9. Added postmaster_mailfrom=xxx to callout verification option.
2912 10. Added mailfrom=xxxx to the callout verification option, for verify=
2915 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2916 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2918 12. Inserted some casts to stop certain compilers warning when using pointer
2919 differences as field lengths or precisions in printf-type calls (mostly
2920 affecting debugging statements).
2922 13. Added optional readline() support for -be (dynamically loaded).
2924 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2925 same clock tick as a message's arrival, so that its received time was the
2926 same as the "first fail" time on the retry record, and that message
2927 remained on the queue past the ultimate address timeout, every queue runner
2928 would try a delivery (because it was past the ultimate address timeout) but
2929 after another failure, the ultimate address timeout, which should have then
2930 bounced the address, did not kick in. This was a "< instead of <=" error;
2931 in most cases the first failure would have been in the next clock tick
2932 after the received time, and all would be well.
2934 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2935 being recognized when the domain list was tested by the match_domain
2936 condition in an expansion string.
2938 16. Added the ${str2b64: operator.
2940 17. Exim was always calling setrlimit() to set a large limit for the number of
2941 processes, without checking whether the existing limit was already
2942 adequate. (It did check for the limit on file descriptors.) Furthermore,
2943 errors from getrlimit() and setrlimit() were being ignored. Now they are
2944 logged to the main and panic logs, but Exim does carry on, to try to do its
2945 job under whatever limits there are.
2947 18. Imported PCRE 5.0.
2949 19. Trivial typo in log message " temporarily refused connection" (the leading
2952 20. If the log selector return_path_on_delivery was set and an address was
2953 redirected to /dev/null, the delivery process crashed because it assumed
2954 that a return path would always be set for a "successful" delivery. In this
2955 case, the whole delivery is bypassed as an optimization, and therefore no
2958 21. Internal re-arrangement: the function for sending a challenge and reading
2959 a response while authentication was assuming a zero-terminated challenge
2960 string. It's now changed to take a pointer and a length, to allow for
2961 binary data in such strings.
2963 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2965 23. Exim was not respecting finduser_retries when seeking the login of the
2966 uid under which it was called; it was always trying 10 times. (The default
2967 setting of finduser_retries is zero.) Also, it was sleeping after the final
2968 failure, which is pointless.
2970 24. Implemented tls_on_connect_ports.
2972 25. Implemented acl_smtp_predata.
2974 26. If the domain in control=submission is set empty, Exim assumes that the
2975 authenticated id is a complete email address when it generates From: or
2976 Sender: header lines.
2978 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2979 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2980 chown and chgrp in /bin and hostname in /usr/bin.
2982 28. Exim was keeping the "process log" file open after each use, just as it
2983 does for the main log. This opens the possibility of it remaining open for
2984 long periods when the USR1 signal hits a daemon. Occasional processlog
2985 errors were reported, that could have been caused by this. Anyway, it seems
2986 much more sensible not to leave this file open at all, so that is what now
2989 29. The long-running daemon process does not normally write to the log once it
2990 has entered its main loop, and it closes the log before doing so. This is
2991 so that log files can straightforwardly be renamed and moved. However,
2992 there are a couple of unusual error situations where the daemon does write
2993 log entries, and I had neglected to close the log afterwards.
2995 30. The text of an SMTP error response that was received during a remote
2996 delivery was being truncated at 512 bytes. This is too short for some of
2997 the long messages that one sometimes sees. I've increased the limit to
3000 31. It is now possible to make retry rules that apply only when a message has a
3001 specific sender, in particular, an empty sender.
3003 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3004 it possible to be selective about when SMTP synchronization is enforced.
3006 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3008 32. Implemented hosts_connection_nolog.
3010 33. Added an ACL for QUIT.
3012 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3015 35. Added mailbox_size and mailbox_filecount to appendfile.
3017 36. Added control = no_multiline_responses to ACLs.
3019 37. There was a bug in the logic of the code that waits for the clock to tick
3020 in the case where the clock went backwards by a substantial amount such
3021 that the microsecond fraction of "now" was more than the microsecond
3022 fraction of "then" (but the whole seconds number was less).
3024 38. Added support for the libradius Radius client library this is found on
3025 FreeBSD (previously only the radiusclient library was supported).
3031 1. When certain lookups returned multiple values in the form name=value, the
3032 quoting of the values was not always being done properly. Specifically:
3033 (a) If the value started with a double quote, but contained no whitespace,
3035 (b) If the value contained whitespace other than a space character (i.e.
3036 tabs or newlines or carriage returns) it was not quoted.
3037 This fix has been applied to the mysql and pgsql lookups by writing a
3038 separate quoting function and calling it from the lookup code. The fix
3039 should probably also be applied to nisplus, ibase and oracle lookups, but
3040 since I cannot test any of those, I have not disturbed their existing code.
3042 2. A hit in the callout cache for a specific address caused a log line with no
3043 reason for rejecting RCPT. Now it says "Previous (cached) callout
3044 verification failure".
3046 3. There was an off-by-one bug in the queryprogram router. An over-long
3047 return line was truncated at 256 instead of 255 characters, thereby
3048 overflowing its buffer with the terminating zero. As well as fixing this, I
3049 have increased the buffer size to 1024 (and made a note to document this).
3051 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3052 when Exim is waiting for an SMTP response from a remote server, Exim
3053 restarts its select() call on the socket, thereby resetting its timeout.
3054 This is not a problem when such interrupts are rare. Somebody set up a cron
3055 job to run exiwhat every 2 minutes, which is less than the normal select()
3056 timeout (5 or 10 minutes). This meant that the select() timeout never
3057 kicked in because it was always reset. I have fixed this by comparing the
3058 time when an interrupt arrives with the time at the start of the first call
3059 to select(). If more time than the timeout has elapsed, the interrupt is
3060 treated as a timeout.
3062 5. Some internal re-factoring in preparation for the addition of Sieve
3063 extensions (by MH). In particular, the "personal" test is moved to a
3064 separate function, and given an option for scanning Cc: and Bcc: (which is
3065 not set for Exim filters).
3067 6. When Exim created an email address using the login of the caller as the
3068 local part (e.g. when creating a From: or Sender: header line), it was not
3069 quoting the local part when it contained special characters such as @.
3071 7. Installed new OpenBSD configuration files.
3073 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3074 try to make them clearer.
3076 9. Callout options, other than the timeout value, were being ignored when
3077 verifying sender addresses in header lines. For example, when using
3079 verify = header_sender/callout=no_cache
3081 the cache was (incorrectly) being used.
3083 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3084 only the Cygwin environment.
3086 11. When return_path_on_delivery was set as a log selector, if different remote
3087 addresses in the same message used different return paths and parallel
3088 remote delivery occurred, the wrong values would sometimes be logged.
3089 (Whenever a remote delivery process finished, the return path value from
3090 the most recently started remote delivery process was logged.)
3092 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3093 lines when AUTH and/or TLS are in use. This is the "received protocol"
3094 field. Exim used to use "asmtp" for authenticated SMTP, without any
3095 indication (in the protocol name) for TLS use. Now it follows the RFC and
3096 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3097 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3098 names appear in log lines as well as in Received: header lines.
3100 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3101 extensions, and comparison tests, and to fix some bugs.
3103 14. Changes to the "personal" filter test:
3105 (1) The test was buggy in that it was just doing the equivalent of
3106 "contains" tests on header lines. For example, if a user's address was
3107 anne@some.where, the "personal" test would incorrectly be true for
3109 To: susanne@some.where
3111 This test is now done by extracting each address from the header in turn,
3112 and checking the entire address. Other tests that are part of "personal"
3113 are now done using regular expressions (for example, to check local parts
3114 of addresses in From: header lines).
3116 (2) The list of non-personal local parts in From: addresses has been
3117 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3118 taken from the Sieve specification recommendations.
3120 (3) If the message contains any header line starting with "List-" it is
3121 treated as non-personal.
3123 (4) The test for "circular" in the Subject: header line has been removed
3124 because it now seems ill-conceived.
3126 15. Minor typos in src/EDITME comments corrected.
3128 16. Installed latest exipick from John Jetmore.
3130 17. If headers_add on a router specified a text string that was too long for
3131 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3132 of string_sprintf() is now avoided.
3134 18. $message_body_size was not set (it was always zero) when running the DATA
3135 ACL and the local_scan() function.
3137 19. For the "mail" command in an Exim filter, no default was being set for
3138 the once_repeat time, causing a random time value to be used if "once" was
3139 specified. (If the value happened to be <= 0, no repeat happened.) The
3140 default is now 0s, meaning "never repeat". The "vacation" command was OK
3141 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3142 (I found it when inspecting the code).
3144 20. There is now an overall timeout for performing a callout verification. It
3145 defaults to 4 times the callout timeout, which applies to individual SMTP
3146 commands during the callout. The overall timeout applies when there is more
3147 than one host that can be tried. The timeout is checked before trying the
3148 next host. This prevents very long delays if there are a large number of
3149 hosts and all are timing out (e.g. when the network connections are timing
3150 out). The value of the overall timeout can be changed by specifying an
3151 additional sub-option for "callout", called "maxwait". For example:
3153 verify = sender/callout=5s,maxwait=20s
3155 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3156 to the end before writing, but this should make it even safer).
3158 22. Exim was forgetting that it had advertised PIPELINING for the second and
3159 subsequent messages on an SMTP connection. It was also not resetting its
3160 memory on STARTTLS and an internal HELO.
3162 23. When Exim logs an SMTP synchronization error within a session, it now
3163 records whether PIPELINING has been advertised or not.
3165 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3166 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3167 rather than long int.
3169 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3171 26. Added the never_mail option to autoreply.
3177 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3178 crash if the getsockname() call failed; this can happen if a connection is
3179 closed very soon after it is established. The problem was simply in the
3180 order in which certain operations were done, causing Exim to try to write
3181 to the SMTP stream before it had set up the file descriptor. The bug has
3182 been fixed by making things happen in the correct order.
3188 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3189 before the connection was closed, thus losing the rejection response.
3191 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3192 some early Solaris releases, but causes trouble in current releases where
3193 socklen_t is defined.
3195 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3198 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3201 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3202 that releasing the top of it at the end releases what was used for sub-
3203 expansions (unless the block got too big). However, discard this block if
3204 the first thing is a variable or header, so that we can use its block when
3205 it is dynamic (useful for very large $message_headers, for example).
3207 6. Lookups now cache *every* query, not just the most recent. A new, separate
3208 store pool is used for this. It can be recovered when all lookup caches are
3209 flushed. Lookups now release memory at the end of their result strings.
3210 This has involved some general refactoring of the lookup sources.
3212 7. Some code has been added to the store_xxx() functions to reduce the amount
3213 of flapping under certain conditions.
3215 8. log_incoming_interface used to affect only the <= reception log lines. Now
3216 it causes the local interface and port to be added to several more SMTP log
3217 lines, for example "SMTP connection from", and rejection lines.
3219 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3221 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3223 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3224 could be overwritten at the end of the current message (or the start of a
3225 new message if it was set in a HELO ACL). The value is now preserved for
3226 the duration of the SMTP connection.
3228 12. If a transport had a headers_rewrite setting, and a matching header line
3229 contained an unqualified address, that address was qualified, even if it
3230 did not match any rewriting rules. The underlying bug was that the values
3231 of the flags that permit the existence of unqualified sender and recipient
3232 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3233 non-local messages, and by -bnq for local messages) were not being
3234 preserved with the message after it was received.
3236 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3237 "next input=" as part of the text comprising the host identity instead of
3238 the correct text. The code was using the same buffer for two different
3239 strings. However, depending on which order the printing function evaluated
3240 its arguments, the bug did not always show up. Under Linux, for example, my
3241 test suite worked just fine.
3243 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3244 doesn't work with some older versions of Perl. It has been changed to "my",
3245 which in any case is probably the better facility to use.
3247 15. A really picky compiler found some instances of statements for creating
3248 error messages that either had too many or two few arguments for the format
3251 16. The size of the buffer for calls to the DNS resolver has been increased
3252 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3253 for addresses that have a lot of PTR records. This alleviates a problem; it
3254 does not fully solve it.
3256 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3257 buffer now truncates the list and logs the incident, which is the same
3258 action as happens when Exim is looking up a host name and its aliases.
3259 Previously in this situation something unpredictable would happen;
3260 sometimes it was "internal error: store_reset failed".
3262 18. If a server dropped the connection unexpectedly when an Exim client was
3263 using GnuTLS and trying to read a response, the client delivery process
3264 crashed while trying to generate an error log message.
3266 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3267 string, for example:
3269 warn message = H1: something\nH2: something
3271 the text was added as a single header line from Exim's point of view
3272 though it ended up OK in the delivered message. However, searching for the
3273 second and subsequent header lines using $h_h2: did not work. This has been
3274 fixed. Similarly, if a system filter added multiple headers in this way,
3275 the routers could not see them.
3277 20. Expanded the error message when iplsearch is called with an invalid key to
3278 suggest using net-iplsearch in a host list.
3280 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3281 ACLs are no longer actually imposed (and a message to that effect is
3284 22. If a "gecos" field in a passwd entry contained escaped characters, in
3285 particular, if it contained a \" sequence, Exim got it wrong when building
3286 a From: or a Sender: header from that name. A second bug also caused
3287 incorrect handling when an unquoted " was present following a character
3288 that needed quoting.
3290 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3291 was not being matched caselessly.
3293 24. Arranged for all hyphens in the exim.8 source to be escaped with
3296 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3297 properly. Recipient callout cache records were still being keyed to include
3298 the sender, even when use_sender was set false. This led to far more
3299 callouts that were necessary. The sender is no longer included in the key
3300 when use_sender is false.
3302 26. Added "control = submission" modifier to ACLs.
3304 27. Added the ${base62d: operator to decode base 62 numbers.
3306 28. dnsdb lookups can now access SRV records.
3308 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3309 the configuration file.
3311 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3312 (-v) mode. This makes the output for a verbose queue run more intelligible.
3314 31. Added a use_postmaster feature to recipient callouts.
3316 32. Added the $body_zerocount variable, containing the number of binary zero
3317 bytes in the message body.
3319 33. The time of last modification of the "new" subdirectory is now used as the
3320 "mailbox time last read" when there is a quota error for a maildir
3323 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3325 35. Added +ignore_unknown as a special item in host lists.
3327 36. Code for decoding IPv6 addresses in host lists is now included, even if
3328 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3329 address was recognized as an IP address, but was then not correctly decoded
3330 into binary, causing unexpected and incorrect effects when compared with
3337 1. Very minor rewording of debugging text in manualroute to say "list of
3338 hosts" instead of "hostlist".
3340 2. If verify=header_syntax was set, and a header line with an unqualified
3341 address (no domain) and a large number of spaces between the end of the
3342 name and the colon was received, the reception process suffered a buffer
3343 overflow, and (when I tested it) crashed. This was caused by some obsolete
3344 code that should have been removed. The fix is to remove it!
3346 3. When running in the test harness, delay a bit after writing a bounce
3347 message to get a bit more predictability in the log output.
3349 4. Added a call to search_tidyup() just before forking a reception process. In
3350 theory, someone could use a lookup in the expansion of smtp_accept_max_
3351 per_host which, without the tidyup, could leave open a database connection.
3353 5. Added the variables $recipient_data and $sender_data which get set from a
3354 lookup success in an ACL "recipients" or "senders" condition, or a router
3355 "senders" option, similar to $domain_data and $local_part_data.
3357 6. Moved the writing of debug_print from before to after the "senders" test
3360 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3361 problems for message scanning, either using a data ACL, or using
3362 local_scan() because the Received: header was not generated till after they
3363 were called (in order to set the time as the time of reception completion).
3364 I have revised the way this works. The header is now generated after the
3365 body is received, but before the ACL or local_scan() are called. After they
3366 are run, the timestamp in the header is updated.
3372 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3373 before starting a queue runner without re-exec. This happened only when
3374 deliver_drop_privilege was set or when the Exim user was set to root. The
3375 effect of the bug was that timeouts during subsequent deliveries caused
3376 crashes instead of being properly handled. The handler is now left at its
3377 default (and expected) setting.
3379 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3380 message, again when deliver_drop_privilege is set or Exim is run as root.
3381 The bug described in (1) was not present in this case, but the tidying up
3382 of the other signals was missing. I have made the two cases consistent.
3384 3. The ignore_target_hosts setting on a manualroute router was being ignored
3385 for hosts that were looked up using the /MX notation.
3387 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3390 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3391 operated on the sender address. After changing the $sender_address to <>
3392 for the sender address verify, Exim was re-instated it as the original
3393 (before rewriting) address, but remembering that it had rewritten it, so it
3394 wasn't rewriting it again. This bug also had the effect of breaking the
3395 sender address verification caching when the sender address was rewritten.
3397 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3398 This has been changed so that if the ip literal address matches
3399 ignore_target_hosts, the router declines.
3401 7. Added expansion conditions match_domain, match_address, and match_local_
3402 part (NOT match_host).
3404 8. The placeholder for the Received: header didn't have a length field set.
3406 9. Added code to Exim itself and to exim_lock to test for a specific race
3407 condition that could lead to file corruption when using MBX delivery. The
3408 issue is with the lockfile that is created in /tmp. If this file is removed
3409 after a process has opened it but before that process has acquired a lock,
3410 there is the potential for a second process to recreate the file and also
3411 acquire a lock. This could lead to two Exim processes writing to the file
3412 at the same time. The added code performs the same test as UW imapd; it
3413 checks after acquiring the lock that its file descriptor still refers to
3414 the same named file.
3416 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3417 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3420 11. Added the smtp_active_hostname option. If used, this will typically be made
3421 to depend on the incoming interface address. Because $interface_address is
3422 not set up until the daemon has forked a reception process, error responses
3423 that can happen earlier (such as "too many connections") no longer contain
3426 12. If an expansion in a condition on a "warn" statement fails because a lookup
3427 defers, the "warn" statement is abandoned, and the next ACL statement is
3428 processed. Previously this caused the whole ACL to be aborted.
3430 13. Added the iplsearch lookup type.
3432 14. Added ident_timeout as a log selector.
3434 15. Added tls_certificate_verified as a log selector.
3436 16. Added a global option tls_require_ciphers (compare the smtp transport
3437 option of the same name). This controls incoming TLS connections.
3439 17. I finally figured out how to make tls_require_ciphers do a similar thing
3440 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3441 before starting the TLS session.
3443 18. Tabs are now shown as \t in -bP output.
3445 19. If the log selector return_path_on_delivery was set, Exim crashed when
3446 bouncing a message because it had too many Received: header lines.
3448 20. If two routers both had headers_remove settings, and the first one included
3449 a superfluous trailing colon, the final name in the first list and the
3450 first name in the second list were incorrectly joined into one item (with a
3451 colon in the middle).
3457 1. Added -C and -D options to the exinext utility, mainly to make it easier
3458 to include in the automated testing, but these could be helpful when
3459 multiple configurations are in use.
3461 2. The exinext utility was not formatting the output nicely when there was
3462 an alternate port involved in the retry record key, nor when there was a
3463 message id as well (for retries that were specific to a specific message
3464 and a specific host). It was also confused by IPv6 addresses, because of
3465 the additional colons they contain. I have fixed the IPv4 problem, and
3466 patched it up to do a reasonable job for IPv6.
3468 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3469 delivery, the log line now contains "pipelined" if PIPELINING was used.
3471 4. An SMTP transport process used to panic and die if the bind() call to set
3472 an explicit outgoing interface failed. This has been changed; it is now
3473 treated in the same way as a connect() failure.
3475 5. A reference to $sender_host_name in the part of a conditional expansion
3476 that was being skipped was still causing a DNS lookup. This no longer
3479 6. The def: expansion condition was not recognizing references to header lines
3480 that used bh_ and bheader_.
3482 7. Added the _cache feature to named lists.
3484 8. The code for checking quota_filecount in the appendfile transport was
3485 allowing one more file than it should have been.
3487 9. For compatibility with Sendmail, the command line option
3495 and sets the incoming protocol and host name (for trusted callers). The
3496 host name and its colon can be omitted when only the protocol is to be set.
3497 Note the Exim already has two private options, -pd and -ps, that refer to
3498 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3499 "s", but I don't think that's a major issue.
3501 10. A number of refactoring changes to the code, none of which should affect
3504 (a) The number of logging options was getting close to filling up the
3505 32-bit word that was used as a bit map. I have split them into two classes:
3506 those that are passed in the argument to log_write(), and those that are
3507 only ever tested independently outside of that function. These are now in
3508 separate 32-bit words, so there is plenty of room for expansion again.
3509 There is no change in the user interface or the logging behaviour.
3511 (b) When building, for example, log lines, the code previously used a
3512 macro that called string_cat() twice, in order to add two strings. This is
3513 not really sufficiently general. Furthermore, there was one instance where
3514 it was actually wrong because one of the argument was used twice, and in
3515 one call a function was used. (As it happened, calling the function twice
3516 did not affect the overall behaviour.) The macro has been replaced by a
3517 function that can join an arbitrary number of extra strings onto a growing
3520 (c) The code for expansion conditions now uses a table and a binary chop
3521 instead of a serial search (which was left over from when there were very
3522 few conditions). Also, it now recognizes conditions like "pam" even when
3523 the relevant support is not compiled in: a suitably worded error message is
3524 given if an attempt is made to use such a condition.
3526 11. Added ${time_interval:xxxxx}.
3528 12. A bug was causing one of the ddress fields not to be passed back correctly
3529 from remote delivery subprocesses. The field in question was not being
3530 subsequently used, so this caused to problems in practice.
3532 13. Added new log selectors queue_time and deliver_time.
3534 14. Might have fixed a bug in maildirsizefile handling that threw up
3535 "unexpected character" debug warnings, and recalculated the data
3536 unnecessarily. In any case, I expanded the warning message to give more
3539 15. Added the message "Restricted characters in address" to the statements in
3540 the default ACL that block characters like @ and % in local parts.
3542 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3543 Three changes have been made:
3545 (a) There was a serious bug; a negative response to MAIL caused the whole
3546 recipient domain to be cached as invalid, thereby blocking all messages
3547 to all local parts at the same domain, from all senders. This bug has
3548 been fixed. The domain is no longer cached after a negative response to
3549 MAIL if the sender used is not empty.
3551 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3554 (c) A new callout option, "use_sender" has been added for people who want
3555 the modified behaviour.
3561 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3564 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3565 indeed breaks things for older releases.
3567 3. Added additional logging to the case where there is a problem reading data
3568 from a filter that is running in a subprocess using a pipe, in order to
3569 try to track down a specific problem.
3571 4. Testing facility fudge: when running in the test harness and attempting
3572 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3573 getting "No route to host". Convert this to a timeout.
3575 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3578 6. Some OS don't have socklen_t but use size_t instead. This affects the
3579 fifth argument of getsockopt() amongst other things. This is now
3580 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3581 can be set for individual OS. I have set it for SunOS5, OSF1, and
3582 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3583 some earlier ones do not.
3585 7. Change 4.30/15 was not doing the test caselessly.
3587 8. The standard form for an IPv6 address literal was being rejected by address
3588 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3589 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3590 this, as well as the form without the "IPv6" on the front (but only when
3591 address literals are enabled, of course).
3593 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3595 10. Exim crashed if a message with an empty sender address specified by -f
3596 encountered a router with an errors_to setting. This could be provoked only
3597 by a command such as
3601 where an empty string was supplied; "<>" did not hit this bug.
3603 11. Installed PCRE release 4.5.
3605 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3606 remained set. It is now erased.
3608 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3609 times from message ids (which are base 36 rather than the normal 62).
3611 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3612 were being counted as actual protocol errors, and logged if the log
3613 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3614 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3615 and DATA following a set of rejected RCPTs do not count as protocol errors.
3616 In other words, Exim assumes they were pipelined, though this may not
3617 actually be the case. Of course, in all cases the client gets an
3618 appropriate error code.
3620 15. If a lookup fails in an ACL condition, a message about the failure may
3621 be available; it is used if testing the ACL cannot continue, because most
3622 such messages specify what the cause of the deferral is. However, some
3623 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3624 that caused an old message to be retained and used if a later statement
3625 caused a defer, replacing the real cause of the deferral.
3627 16. If an IP address had so many PTR records that the DNS lookup buffer
3628 was not large enough to hold them, Exim could crash while trying to process
3629 the truncated data. It now detects and logs this case.
3631 17. Further to 4.21/58, another change has been made: if (and only if) the
3632 first line of a message (the first header line) ends with CRLF, a bare LF
3633 in a subsequent header line has a space inserted after it, so as not to
3634 terminate the header.
3636 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3637 unnecessarily, used atoi() instead of strtol(), and didn't check the
3638 termination when getting file sizes from file names by regex.
3640 19. Completely re-implemented the support for maildirsize files, in the light
3641 of a number of problems with the previous contributed implementation
3642 (4.30/29). In particular:
3644 . If the quota is zero, the maildirsize file is maintained, but no quota is
3647 . If the maildir directory does not exist, it is created before any attempt
3648 to write a maildirsize file.
3650 . The quota value in the file is just a cache; if the quota is changed in
3651 the transport, the new value overrides.
3653 . A regular expression is available for excluding directories from the
3656 20. The autoreply transport checks the characters in options that define the
3657 message's headers; it allows continued headers, but it was checking with
3658 isspace() after an embedded newline instead of explicitly looking for a
3661 21. If all the "regular" hosts to which an address was routed had passed their
3662 expiry times, and had not reached their retry times, the address was
3663 bounced, even if fallback hosts were defined. Now Exim should go on to try
3666 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3667 equivalent code in the SMTP transport. Some hosts send humungous responses
3668 to HELO/EHLO, more than 1024 it seems.
3670 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3671 gives compiler warnings in some environments. I've now done it "properly",
3674 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3675 (because of problems with the built-in one) was declared to return uschar *
3676 instead of char *, causing compiler failure.
3678 25. Fixed a file descriptor leak when processing alias/forward files.
3680 26. Fixed a minor format string issue in dbfn.c.
3682 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3684 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3685 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3688 29. When a sender address is verified, it is cached, to save repeating the test
3689 when there is more than one recipient in a message. However, when the
3690 verification involves a callout, it is possible for different callout
3691 options to be set for different recipients. It is too complicated to keep
3692 track of this in the cache, so now Exim always runs a verification when a
3693 callout is required, relying on the callout cache for the optimization.
3694 The overhead is duplication of the address routing, but this should not be
3697 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3698 to be verified with callout=postmaster, and the main callout worked but the
3699 postmaster check failed, the verification correctly failed. However, if a
3700 subsequent RCPT command asked for sender verification *without* the
3701 postmaster check, incorrect caching caused this verification also to fail,
3704 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3705 it was not caching the DNS options (qualify_single, search_parents) that
3706 were used when the lookup failed. A subsequent lookup with different
3707 options therefore always gave the same answer, though there were cases
3708 where it should not have. (Example: a "domains = !$mx_any" option on a
3709 dnslookup router: the "domains" option is always processed without any
3710 widening, but the router might have qualify_single set.) Now Exim uses the
3711 cached value only when the same options are set.
3713 32. Added John Jetmore's "exipick" utility to the distribution.
3715 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3716 than a timeout (e.g. a certificate is required, and is not provided), an
3717 Exim server now closes the connection immediately. Previously it waited for
3718 the client to close - but if the client is SSL, it seems that they each
3719 wait for each other, leading to a delay before one of them times out.
3721 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3722 maintained 0.8.x compatibility because I don't think many are using it, and
3723 it is clearly obsolete.
3725 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3728 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3729 client certificate was expired. A simple patch fixes this, though I don't
3730 understand the full logic of why the verify callback is called multiple
3733 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3734 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3735 which causes problems with some clients (such as the Certicom SSL Plus
3736 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3737 disables the coutermeasure allowing Eudora to connect."
3739 38. Exim was not checking that a write() to a log file succeeded. This could
3740 lead to Bad Things if a log got too big, in particular if it hit a file
3741 size limit. Exim now panics and dies if it cannot write to a log file, just
3742 as it does if it cannot open a log file.
3744 39. Modified OS/Makefile-Linux so that it now contains
3746 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3748 The two -D definitions ensure that Exim is compiled with large file
3749 support, which makes it possible to handle log files that are bigger than
3752 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3753 instance) a domain was checked against a named list that involved a lookup,
3754 causing $domain_data to be set, then another domain was checked against the
3755 same list, then the first domain was re-checked, the value of $domain_data
3756 after the final check could be wrong. In particular, if the second check
3757 failed, it could be set empty. This bug probably also applied to
3760 41. The strip_trailing_dot option was not being applied to the address given
3761 with the -f command-line option.
3763 42. The code for reading a message's header from the spool was incrementing
3764 $received_count, but never initializing it. This meant that the value was
3765 incorrect (doubled) while delivering a message in the same process in which
3766 it was received. In the most common configuration of Exim, this never
3767 happens - a fresh exec is done - but it can happen when
3768 deliver_drop_privilege is set.
3770 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3771 it now includes up to 150 characters of the unexpected data in the log
3774 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3775 and building data strings. The size of both of these buffers was 10 000
3776 bytes - far larger than anybody would *ever* want, thought I. Needless to
3777 say, somebody hit the limit. I have increased the maximum line length to
3778 20 000 and the maximum data length of concatenated lines to 100 000. I have
3779 also fixed two bugs, because there was no checking on these buffers. Tsk,
3780 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3781 buffer is too small.
3783 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3784 lsearch lookups. Now it does.
3786 46. When parsing a route_list item in a manualroute router, a fixed-length
3787 buffer was used for the list of hosts. I made this 1024 bytes long,
3788 thinking that nobody would ever have a list of hosts that long. Wrong.
3789 Somebody had a whole pile of complicated expansion conditions, and the
3790 string was silently truncated, leading to an expansion error. It turns out
3791 that it is easier to change to an unlimited length (owing to other changes
3792 that have happened since this code was originally written) than to build
3793 structure for giving a limitation error. The length of the item that
3794 expands into the list of hosts is now unlimited.
3796 47. The lsearch lookup could not handle data where the length of text line was
3797 more than 4095 characters. Such lines were truncated, leading to shortened
3798 data being returned. It should now handle lines of any length.
3800 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3801 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3804 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3805 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3806 no longer happens in comment lines. A list of replacements is now placed
3807 at the head of all of the source files, except those whose only change is
3808 to replace PERL_COMMAND in the very first #! line.
3810 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3811 messages on the queue, with a bottom-up merge sort, using code contributed
3812 by Michael Haardt. This should make operations like -bp somewhat faster on
3813 large queues. It won't affect queue runners, except when queue_run_in_order
3816 51. Installed eximstats 1.31 in the distribution.
3818 52. Added support for SRV lookups to the dnslookup router.
3820 53. If an ACL referred to $message_body or $message_body_end, the value was not
3821 reset for any messages that followed in the same SMTP session.
3823 54. The store-handling optimization for building very long strings was not
3824 differentiating between the different store pools. I don't think this
3825 actually made any difference in practice, but I've tidied it.
3827 55. While running the routers to verify a sender address, $sender_address
3828 was still set to the sender address. This is wrong, because when routing to
3829 send a bounce to the sender, it would be empty. Therefore, I have changed
3830 it so that, while verifying a sender address, $sender_address is set to <>.
3831 (There is no change to what happens when verifying a recipient address.)
3833 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3834 target A or AAAA records (if not already returned) without resetting the
3835 qualify_single or search_parents options of the DNS resolver. These are
3836 inappropriate in this case because the targets of MX and SRV records must
3837 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3838 target that, when qualified, matched something in the local domain. These
3839 two options are now turned off when doing these lookups.
3841 57. It seems that at least some releases of Reiserfs (which does not have the
3842 concept of a fixed number of inodes) returns zero and not -1 for the
3843 number of available inodes. This interacted badly with check_spool_inodes,
3844 which assumed that -1 was the "no such thing" setting. What I have done is
3845 to check that the total number of inodes is greater than zero before doing
3846 the test of how many are available.
3848 58. When a "warn" ACL statement has a log_message modifier, the message is
3849 remembered, and not repeated. This is to avoid a lot of repetition when a
3850 message has many recipients that cause the same warning to be written.
3851 Howewer, Exim was preserving the list of already written lines for an
3852 entire SMTP session, which doesn't seem right. The memory is now reset if a
3853 new message is started.
3855 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3856 debugging output unless log_rewrite was also set.
3858 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3859 of (int)(handle) when we know that handle contains (void *)(-1).
3861 61. The Exim daemon panic-logs an error return when it closes the incoming
3862 connection. However "connection reset by peer" seems to be common, and
3863 isn't really an error worthy of noting specially, so that particular error
3866 62. When Exim is trying to find all the local interfaces, it used to panic and
3867 die if the ioctl to get the interface flags failed. However, it seems that
3868 on at least one OS (Solaris 9) it is possible to have an interface that is
3869 included in the list of interfaces, but for which you get a failure error
3870 for this call. This happens when the interface is not "plumbed" into a
3871 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3872 failure of the "get flags" call assumes that the interface is down.
3874 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3875 makes life easier for people who are doing arithmetic on fields extracted
3876 from dates, where you often get leading zeros that should not be
3877 interpreted as octal.
3879 64. Added qualify_domain to the redirect router, to override the global
3882 65. If a pathologically long header line contained very many addresses (the
3883 report of this problem mentioned 10 000) and each of them was rewritten,
3884 Exim could use up a very large amount of memory. (It kept on making new
3885 copies of the header line as it rewrote, and never released the old ones.)
3886 At the expense of a bit more processing, the header rewriting function has
3887 been changed so that it no longer eats memory in this way.
3889 66. The generation of the Received: header has been moved from the time that a
3890 message starts to be received, to the time that it finishes. The timestamp
3891 in the Received: header should now be very close to that of the <= log
3892 line. There are two side-effects of this change:
3894 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3895 logged header lines no longer include the local Received: line, because
3896 it has not yet been created. The same applies to a copy of the message
3897 that is returned to a non-SMTP sender when a message is rejected.
3899 (b) When a filter file is tested using -bf, no additional Received: header
3900 is added to the test message. After some thought, I decided that this
3903 This change does not affect the value of $received_for. It is still set
3904 after address rewriting, but before local_scan() is called.
3906 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3908 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3909 gave an unhelpful panic error message, and a defer error. I have managed to
3910 change this behaviour so that it now rejects any supplied certificate,
3911 which seems right, as the list of acceptable certificates is empty.
3913 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3914 gave an unhelpful defer error. I have not managed to make this reject any
3915 supplied certificates, but the error message it gives is "no certificate
3916 supplied", which is not helpful.
3918 70. exigrep's output now also includes lines that are not associated with any
3919 message, but which match the given pattern. Implemented by a patch from
3920 Martin Sluka, which also tidied up the Perl a bit.
3922 71. Recipient callout verification, like sender verification, was using <> in
3923 the MAIL FROM command. This isn't really the right thing, since the actual
3924 sender may affect whether the remote host accepts the recipient or not. I
3925 have changed it to use the actual sender in the callout; this means that
3926 the cache record is now keyed on a recipient/sender pair, not just the
3927 recipient address. There doesn't seem to be a real danger of callout loops,
3928 since a callout by the remote host to check the sender would use <>.
3929 [SEE ABOVE: changed after hitting problems.]
3931 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3932 temporary errors. However, in the case of such a code being given after
3933 the end of a data transmission (i.e. after ".") Exim was failing to write
3934 a retry record for the message. (Yes, there was some broken host that was
3935 actually sending 8xx at this point.)
3937 73. An unknown lookup type in a host list could cause Exim to panic-die when
3938 the list was checked. (An example that provoked this was putting <; in the
3939 middle of a list instead of at the start.) If this happened during a DATA
3940 ACL check, a -D file could be left lying around. This kind of configuration
3941 error no longer causes Exim to die; instead it causes a defer errror. The
3942 incident is still logged to the main and panic logs.
3944 74. Buglet left over from Exim 3 conversion. The message "too many messages
3945 in one connection" was written to the rejectlog but not the mainlog, except
3946 when address rewriting (yes!) was being logged.
3948 75. Added write_rejectlog option.
3950 76. When a system filter was run not as root (that is, when system_filter_user
3951 was set), the values of the $n variables were not being returned to the
3952 main process; thus, they were not subsequently available in the $sn
3955 77. Added +return_path_on_delivery log selector.
3957 78. A connection timeout was being treated differently from recipients deferred
3958 when testing hosts_max_try with a message that was older than the host's
3959 retry timeout. (The host should not be counted, thus allowing all hosts to
3960 be tried at least once before bouncing.) This may have been the cause of an
3961 occasionally reported bug whereby a message would remain on the queue
3962 longer than the retry timeout, but would be bounced if a delivery was
3963 forced. I say "may" because I never totally pinned down the problem;
3964 setting up timeout/retry tests is difficult. See also the next item.
3966 79. The ultimate address timeout was not being applied to errors that involved
3967 a combination of host plus message (for example, a timeout on a MAIL
3968 command). When an address resolved to a number of possible hosts, and they
3969 were not all tried for each delivery (e.g. because of hosts_max_try), a
3970 message could remain on the queue longer than the retry timeout.
3972 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3975 81. Fixed an obscure SMTP outgoing bug which required at least the following
3976 conditions: (a) there was another message waiting for the same server;
3977 (b) the server returned 5xx to all RCPT commands in the first message so
3978 that the message was not completed; (c) the server dropped the connection
3979 or gave a negative response to the RSET that Exim sends to abort the
3980 transaction. The observed case was a dropped connection after DATA that had
3981 been sent in pipelining mode. That is, the server had advertised PIPELINING
3982 but was not implementing it correctly. The effect of the bug was incorrect
3983 behaviour, such as trying another host, and this could lead to a crash.
3989 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3990 and daemon.c were passed as pointers to ints; they should have been
3991 pointers to socklen_t variables (which are typically unsigned ints).
3993 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3996 3. Fixed a really odd bug that affected only the testing scheme; patching a
3997 certain fixed string in the binary changed the value of another string that
3998 happened to be identical to the end of the original first string.
4000 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4001 name", it returns that address as the IP address. On some operating
4002 systems (e.g. Solaris), it also passes back the IP address string as the
4003 "host name". However, on others (e.g. Linux), it passes back an empty
4004 string. Exim wasn't checking for this, and was changing the host name to an
4005 empty string, assuming it had been canonicized.
4007 5. Although rare, it is permitted to have more than one PTR record for a given
4008 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4009 all the names associated with an address, because they do in Solaris.
4010 However, it seems that they do not in Linux for data that comes from the
4011 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4012 I found this out when I moved to a new Linux workstation and tried to run
4013 the Exim test suite.
4015 To get round this problem I have changed the code so that it now does its
4016 own call to the DNS to look up PTR records when searching for a host name.
4017 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4018 addresses that are only in /etc/hosts are still found.
4020 This behaviour is, however, controlled by an option called host_lookup_
4021 order, which defaults to "bydns:byaddr". If people want to use the other
4022 order, or indeed, just use one or the other means of lookup, they can
4023 specify it in this variable.
4025 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4026 some operating systems, this comes back from gethostbyaddr() as an empty
4027 string, and this is what Exim used to test for. However, it seems that in
4028 other systems, "." is yielded. Exim now tests for this case too.
4030 7. The values of check_spool_space and check_log_space are now held internally
4031 as a number of kilobytes instead of an absolute number of bytes. If a
4032 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4033 kilobyte. This means that much larger values can be stored.
4035 8. Exim monitor: an attempt to get the action menu when not actually pointing
4036 at a message produces an empty menu entitled "No message selected". This
4037 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4038 no entries in it ("Shell widget menu has zero width and/or height"). So I
4039 have added a single, blank menu entry in this case.
4041 9. Added ${quote_local_part.
4043 10. MIME decoding is now applied to the contents of Subject: header lines when
4046 11. Now that a reference to $sender_host_address automatically causes a reverse
4047 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4048 host lookup before query-style lookups in lists that might use this
4049 variable. This has therefore been abolished, and the "net-" prefix is no
4050 longer necessary for query-style lookups.
4052 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4053 have been a typo for LFLAGS, so it has been changed.
4055 13. The install script calls Exim with "-C /dev/null" in order to find the
4056 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4057 to be output. Howeve, since Exim outputs its version number before the
4058 error, it didn't break the script. It just looked ugly. I fixed this by
4059 always allowing "-C /dev/null" if the caller is root.
4061 14. Ignore overlarge ACL variable number when reading spool file - insurance
4062 against a later release with more variables having written the file.
4064 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4065 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4066 this, as well as the form without the "IPv6" on the front.
4068 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4069 OS/Makefile-Darwin file.
4071 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4072 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4074 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4075 variable. However, some OS use ints for inodes. I've added cast to long int
4076 to get rid of the compiler warning.
4078 19. I had forgotten to lock out "/../" in configuration file names when
4079 ALT_CONFIG_PREFIX was set.
4081 20. Routers used for verification do not need to specify transports. However,
4082 if such a router generated a host list, and callout was configured, Exim
4083 crashed, because it could not find a port number from the (non-existent)
4084 transport. It now assumes port 25 in this circumstance.
4086 21. Added the -t option to exigrep.
4088 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4089 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4090 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4091 them is set, LOOKUP_LSEARCH is forced.
4093 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4095 24. Added sender and host information to the "rejected by local_scan()" log
4096 line; previously there was no indication of these.
4098 25. Added .include_if_exists.
4100 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4101 Linux. It turns out that not all file systems support this. Apparently some
4102 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4103 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4104 not supported on the file descriptor, is now ignored when Exim is trying to
4105 sync a directory. This applies only to Linux.
4107 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4109 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4110 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4113 29. Added support for maildirsize files from supplied patch (modified a bit).
4115 30. The use of :fail: followed by an empty string could lead Exim to respond to
4116 sender verification failures with (e.g.):
4118 550 Verification failed for <xxx>
4119 550 Sender verify failed
4121 where the first response line was missing the '-' that indicates it is not
4122 the final line of the response.
4124 31. The loop for finding the name of the user that called Exim had a hardwired
4125 limit of 10; it now uses the value of finduser_retries, which is used for
4126 all other user lookups.
4128 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4131 33. Exim was neglecting to zero errno before one call of strtol() when
4132 expanding a string and expecting an integer value. On some systems this
4133 resulted in spurious "integer overflow" errors. Also, it was casting the
4134 result into an int without checking.
4136 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4137 did not work. The code looks as if it has *never* worked, though it appears
4138 to have been documented since at least releast 1.62. I have made it work.
4140 35. The "timeout_DNS" error in retry rules, also documented since at least
4141 1.62, also never worked. As it isn't clear exactly what this means, and
4142 clearly it isn't a major issue, I have abolished the feature by treating it
4143 as "timeout", and writing a warning to the main and panic logs.
4145 36. The display of retry rules for -brt wasn't always showing the error code
4148 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4149 timeout_connect_A, timeout_connect_MX.
4151 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4152 to the empty sender.
4154 39. The daemon was not analysing the content of -oX till after it had closed
4155 stderr and disconnected from the controlling terminal. This meant that any
4156 syntax errors were only noted on the panic log, and the return code from
4157 the command was 0. By re-arranging the code a little, I've made the
4158 decoding happen first, so such errors now appear on stderr, and the return
4159 code is 1. However, the actual setting up of the sockets still happens in
4160 the disconnected process, so errors there are still only recorded on the
4163 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4164 connections (as happens on some IP stacks) was logged at start up time as
4165 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4166 it from "IPv6 and IPv4", which means that two separate sockets are being
4169 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4170 says whether AF_INET or AF_INET6 was passed as an argument.
4172 42. Exiwhat output was messed up when time zones were included in log
4175 43. Exiwhat now gives more information about the daemon's listening ports,
4176 and whether -tls-on-connect was used.
4178 44. The "port" option of the smtp transport is now expanded.
4180 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4181 silently ignored. Now an error message is written to the main and panic
4184 46. There's a new ACL modifier called "logwrite" which writes to a log file
4185 as soon as it is encountered.
4187 47. Added $local_user_uid and $local_user_gid at routing time.
4189 48. Exim crashed when trying to verify a sender address that was being
4192 49. Exim was recognizing only a space character after ".include". It now also
4193 recognizes a tab character.
4195 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4196 extracting the relevant information from the specification. The man page no
4197 longer contains scrambled data for the -d option, and I've added a section
4198 at the front about calling Exim under different names.
4200 51. Added "extra_headers" argument to the "mail" command in filter files.
4202 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4205 53. Installed eximstats 1.29.
4207 54. Added transport_filter_timeout as a generic transport option.
4209 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4210 Cc: header lines. This was required by RFC 822, but it not required by RFC
4213 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4214 incoming messages that did not have them. Now it does so only if the
4215 message originates locally, that is, if there is no associated remote host
4216 address. When Resent- header lines are present, this applies to the Resent-
4217 lines rather than the non-Resent- lines.
4219 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4220 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4222 58. Messages for configuration errors now include the name of the main
4223 configuration file - useful now that there may be more than one file in a
4224 list (.included file names were always shown).
4226 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4227 for those rare installations that do not start the daemon as root or run it
4228 setuid root. I've cut out the call to initgroups() if the daemon is not
4231 60. The Exim user and group can now be bound into the binary as text strings
4232 that are looked up at the start of Exim's processing.
4234 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4236 62. Added $mailstore_basename variable.
4238 63. Installed patch to sieve.c from Michael Haardt.
4240 64. When Exim failed to open the panic log after failing to open the main log,
4241 the original message it was trying to log was written to stderr and debug
4242 output, but if they were not available (the usual case in production), it
4243 was lost. Now it is written to syslog before the two lines that record the
4244 failures to open the logs.
4246 65. Users' Exim filters run in subprocesses under the user's uid. It is
4247 possible for a "deliver" command or an alias in a "personal" command to
4248 provoke an address rewrite. If logging of address rewriting is configured,
4249 this fails because the process is not running as root or exim. There may be
4250 a better way of dealing with this, but for the moment (because 4.30 needs
4251 to be released), I have disabled address rewrite logging when running a
4252 filter in a non-root, non-exim process.
4258 1. The buildconfig auxiliary program wasn't quoting the value set for
4259 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4260 not defined. This bug was masked in 4.22 by the effect that was fixed in
4263 2. Some messages that were rejected after a message id was allocated were
4264 shown as "incomplete" by exigrep. It no longer does this for messages that
4265 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4267 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4268 have allow_domain_literals set, the ID did not get logged in the <= line.
4269 Domain literals are now always recognized in Message-ID: header lines.
4271 4. The first argument for a ${extract expansion item is the key name or field
4272 number. Leading and trailing spaces in this item were not being ignored,
4273 causing some misleading effects.
4275 5. When deliver_drop_privilege was set, single queue runner processes started
4276 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4277 same command in the process it spins off) were not dropping privilege.
4279 6. When the daemon running as "exim" started a queue runner, it always
4280 re-executed Exim in the spun-off process. This is a waste of effort when
4281 deliver_drop_privilege is set. The new process now just calls the
4282 queue-runner function directly.
4288 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4291 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4292 ignored. Though the use of -oP was forcing the writing of a pid file, it
4293 was always written to the default place.
4295 3. If the message "no IP address found for host xxxx" is generated during
4296 incoming verification, it is now followed by identification of the incoming
4297 connection (so you can more easily find what provoked it).
4299 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4301 5. Added some features to "harden" Exim a bit more against certain attacks:
4303 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4304 be put in Local/Makefile. This is like the never_users runtime option,
4305 but it cannot be overridden. The default setting is "root".
4307 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4308 prefix string with which any file named in a -C command line option
4311 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4312 is retained for -C and -D only if the caller of Exim is root. Without
4313 it, the exim user may also use -C and -D and retain privilege.
4315 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4316 command line option is disabled.
4318 6. Macro names set by the -D option must start with an upper case letter, just
4319 like macro names defined in the configuration file.
4321 7. Added "dereference=" facility to LDAP.
4323 8. Two instances of the typo "uknown" in the source files are fixed.
4325 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4326 the Configure-Makefile script screwed up while processing it.
4328 10. Incorporated PCRE 4.4.
4330 11. The SMTP synchronization check was not operating right at the start of an
4331 SMTP session. For example, it could not catch a HELO sent before the client
4332 waited for the greeting. There is now a check for outstanding input at the
4333 point when the greeting is written. Because of the duplex, asynchronous
4334 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4335 way, but not yet received, when the check is performed.
4337 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4338 on TCP/IP sockets, because this apparently causes some broken clients to
4341 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4342 unchanged) from the Cygwin maintainer.
4344 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4345 was supported instead of testing for mbx. Effectively a typo.
4347 15. The spa authenticator server code was not checking that the input it
4348 received was valid base64.
4350 16. The debug output line for the "set" modifier in ACLs was not showing the
4351 name of the variable that was being set.
4353 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4355 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4356 lookup on its own. Something else was needed to trigger the lookup. For
4357 example, a match in host_lookup or the need for a host name in a host list.
4358 Now, if $sender_host_name is referenced and the host name has not yet been
4359 looked up, a lookup is performed. If the lookup fails, the variable remains
4360 empty, and $host_lookup_failed is set to "1".
4362 19. Added "eqi" as a case-independent comparison operator.
4364 20. The saslauthd authentication condition could segfault if neither service
4365 nor realm was specified.
4367 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4368 error message that was logged was misleading, and incoming SMTP
4369 connections were dropped. The message is now more accurate, and temporary
4370 errors are given to SMTP connections.
4372 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4373 (except RSET) with a 421 error, until QUIT is received. However, it was
4374 failing to send a response to QUIT.
4376 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4377 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4378 ACL is now run after the helo_try_verify_hosts code.
4380 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4381 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4382 case-independent, so other case variants are also recognized). Apparently
4383 some systems use these upper case variants.
4385 25. If more than two messages were waiting for the same host, and a transport
4386 filter was specified for the transport, Exim sent two messages over the
4387 same TCP/IP connection, and then failed with "socket operation on non-
4388 socket" when it tried to send the third.
4390 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4392 27. The extern definition of crypt16() in expand.c was not being excluded when
4393 the OS had its own crypt16() function.
4395 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4396 as a preferred synonym for return_size_limit, both as an option and as an
4399 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4401 30. Changed the default configuration ACL to relax the local part checking rule
4402 for addresses that are not in any local domains. For these addresses,
4403 slashes and pipe symbols are allowed within local parts, but the sequence
4404 /../ is explicitly forbidden.
4406 31. SPA server authentication was not clearing the challenge buffer before
4409 32. log_message in a "warn" ACL statement was writing to the reject log as
4410 well as to the main log, which contradicts the documentation and doesn't
4411 seem right (because no rejection is happening). So I have stopped it.
4413 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4414 However, I am unable to do any testing of this.
4416 34. Fixed an infelicity in the appendfile transport. When checking directories
4417 for a mailbox, to see if any needed to be created, it was accidentally
4418 using path names with one or more superfluous leading slashes; tracing
4419 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4421 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4422 contents are added to the log line that is written for every discarded
4423 recipient. (Previously a log_message setting was ignored.)
4425 36. The ${quote: operator now quotes the string if it is empty.
4427 37. The install script runs exim in order to find its version number. If for
4428 some reason other than non-existence or emptiness, which it checks, it
4429 could not run './exim', it was installing it with an empty version number,
4430 i.e. as "exim-". This error state is now caught, and the installation is
4433 38. An argument was missing from the function that creates an error message
4434 when Exim fails to connect to the socket for saslauthd authentication.
4435 This could cause Exim to crash, or give a corrupted message.
4437 39. Added isip, isip4, and isip6 to ${if conditions.
4439 40. The ACL variables $acl_xx are now saved with the message, and can be
4440 accessed later in routers, transports, and filters.
4442 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4443 strings in the file are not string-expanded.
4445 42. If a MAIL command specified a SIZE value that was too large to fit into an
4446 int variable, the check against message_size_limit failed. Such values are
4447 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4448 day this will have to be increased, but I don't think I want to be around
4449 when emails are that large.
4456 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4457 iconv() is not standard in FreeBSD.
4459 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4460 IPv6 enabled. The observed symptom was a segmentation fault on return from
4461 the function os_common_find_running_interfaces() in src/os.c.
4463 3. In the check_special_case() function in daemon.c I had used "errno" as an
4464 argument name, which causes warnings on some systems. This was basically a
4465 typo, since it was named "eno" in the comments!
4467 4. The code that waits for the clock to tick (at a resolution of some fraction
4468 of a second) so as to ensure message-id uniqueness was always waiting for
4469 at least one whole tick, when it could have waited for less. [This is
4470 almost certainly not relevant at current processor speeds, where it is
4471 unlikely to ever wait at all. But we try to future-proof.]
4473 5. The function that sleeps for a time interval that includes fractions of a
4474 second contained a race. It did not block SIGALRM between setting the
4475 timer, and suspending (a couple of lines later). If the interval was short
4476 and the sigsuspend() was delayed until after it had expired, the suspension
4477 never ended. On busy systems this could lead to processes getting stuck for
4480 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4481 process, before it forks any delivery processes. The open lookup caching
4482 mechanism meant that the open file or database connection was passed into
4483 the delivery process. The problem was that delivery processes always tidy
4484 up cached lookup data. This could cause a problem for the next delivery
4485 process started by the queue runner, because the external queue runner
4486 process does not know about the closure. So the next delivery process
4487 still has data in the lookup cache. In the case of a file lookup, there was
4488 no problem because closing a file descriptor in a subprocess doesn't affect
4489 the parent. However, if the lookup was caching a connection to a database,
4490 the connection was closed, and the second delivery process was likely to
4491 see errors such as "PGSQL: query failed: server closed the connection
4492 unexpectedly". The problem has been fixed by closing all cached lookups
4493 in a queue runner before running a delivery process.
4495 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4496 seem to have the "const" qualifier which it has on other OS. I've
4499 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4500 *of the same type* (client or server) with the same public name that an
4501 error should be diagnosed.
4503 9. When Exim looked up a host name for an IP address, but failed to find the
4504 original IP address when looking up the host name (a safety check), it
4505 output the message "<ip address> does not match any IP for NULL", which was
4506 confusing, to say the least. The bug was that the host name should have
4507 appeared instead of "NULL".
4509 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4510 user that is built into the binary, and the -C or -D options is used, root
4511 privilege is dropped before the configuration file is read. In addition,
4512 logging is switched to stderr instead of the normal log files. If the
4513 configuration then re-defines the Exim user, the unprivileged environment
4514 is probably not what is expected, so Exim logs a panic warning message (but
4517 However, if deliver_drop_privilege is set, the unprivileged state may well
4518 be exactly what is intended, so the warning has been cut out in that case,
4519 and Exim is allowed to try to write to its normal log files.
4525 1. smtp_return_error_details was not giving details for temporary sender
4526 or receiver verification errors.
4528 2. Diagnose a configuration error if two authenticators have the same public
4531 3. Exim used not to create the message log file for a message until the first
4532 delivery attempt. This could be confusing when incoming messages were held
4533 for policy or load reasons. The message log file is now created at the time
4534 the message is received, and an initial "Received" line is written to it.
4536 4. The automatically generated man page for command line options had a minor
4537 bug that caused no ill effects; however, a more serious problem was that
4538 the procedure for building the man page automatically didn't always
4539 operate. Consequently, release 4.20 contains an out-of-date version. This
4540 shouldn't happen again.
4542 5. When building Exim with embedded Perl support, the script that builds the
4543 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4544 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4546 6. The freeze_tell option was not being used for messages that were frozen on
4547 arrival, either by an ACL or by local_scan().
4549 7. Added the smtp_incomplete_transaction log selector.
4551 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4552 was accepting AUTH without a new EHLO.
4554 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4555 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4558 10. Logging of TCP/IP connections (when configured) now happens in the main
4559 daemon process instead of the child process, so that the TCP/IP connection
4560 count is more accurate (but it can never be perfect).
4562 11. The use of "drop" in a nested ACL was not being handled correctly in the
4563 outer ACL. Now, if condition failure induced by the nested "drop" causes
4564 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4565 or "require"), the connection is dropped.
4567 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4568 that yield "discard" can now be used with an "accept" or a "discard" verb,
4569 but an error is generated for any others (because I can't see a useful way
4570 to define what should happen).
4572 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4573 are now processed in the same way as lines in the Exim configuration file.
4574 In particular, continuation lines are supported.
4576 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4578 15. Added -ti meaning -t -i.
4580 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4581 domains, and warn by logging if others are found.
4583 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4584 returned from GIFCONF when Exim is trying to find the list of interfaces on
4585 a host. The code in os.c has been modified to copy each ifreq to an aligned
4586 structure in all cases.
4588 Also, in some cases, the returned ifreq's were being copied to a 'struct
4589 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4590 means the last couple of bytes of an IPv6 address could be chopped if the
4591 ifreq contained only a normal sockaddr (14 bytes storage).
4593 18. Named domain lists were not supported in the hosts_treat_as_local option.
4594 An entry such as +xxxx was not recognized, and was treated as a literal
4597 19. Ensure that header lines added by a DATA ACL are included in the reject log
4598 if the ACL subsequently rejects the message.
4600 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4601 MD5 (which is deprecated).
4603 21. When testing a filter file using -bf, Exim was writing a message when it
4604 took the sender from a "From " line in the message, but it was not doing so
4605 when it took $return_path from a Return-Path: header line. It now does.
4607 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4608 with a valid header line field name (a series of printing characters
4609 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4611 23. Changed "disc" in the source to "disk" to conform to the documentation and
4612 the book and for uniformity.
4614 24. Ignore Sendmail's -Ooption=value command line item.
4616 25. When execve() failed while trying to run a command in a pipe transport,
4617 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4618 could be confused with a return value of 69 from the command itself. This
4619 has been changed to 127, the value the shell returns if it is asked to run
4620 a non-existent command. The wording for the related log line suggests a
4621 non-existent command as the problem.
4623 26. If received_header_text expands to an empty string, do not add a Received:
4624 header line to the message. (Well, it adds a token one on the spool, but
4625 marks it "old" so that it doesn't get used or transmitted.)
4627 27. Installed eximstats 1.28 (addition of -nt option).
4629 28. There was no check for failure on the call to getsockname() in the daemon
4630 code. This can fail if there is a shortage of resources on the system, with
4631 ENOMEM, for example. A temporary error is now given on failure.
4633 29. Contrary to the C standard, it seems that in some environments, the
4634 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4635 program. Exim now does this explicitly; it affects the formatting of
4636 timestamps using strftime().
4638 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4639 complaints. I've now initialized all the variables, to avoid this.
4641 32. Header lines added by a system filter were not being "seen" during
4642 transport-time rewrites.
4644 33. The info_callback() function passed to OpenSSL is set up with type void
4645 (*)(SSL *, int, int), as described somewhere. However, when calling the
4646 function (actually a macro) that sets it up, the type void(*)() is
4647 expected. I've put in a cast to prevent warnings from picky compilers.
4649 34. If a DNS black list lookup found a CNAME record, but there were no A
4650 records associated with the domain it pointed at, Exim crashed.
4652 35. If a DNS black list lookup returned more than one A record, Exim ignored
4653 all but the first. It now scans all returned addresses if a particular IP
4654 value is being sought. In this situation, the contents of the
4655 $dnslist_value variable are a list of all the addresses, separated by a
4658 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4659 to accept a host name and all its aliases if the forward lookup for any of
4660 them yielded the IP address of the incoming connection. Now it accepts only
4661 those names whose forward lookup yields the correct IP address. Any other
4662 names are discarded. This closes a loophole whereby a rogue DNS
4663 administrator could create reverse DNS records to break through a
4664 wildcarded host restriction in an ACL.
4666 37. If a user filter or a system filter that ran in a subprocess used any of
4667 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4668 the wrong values were passed to the pipe command ($thisaddress had the
4669 value of $0, $0 had the value of $1, etc). This bug was introduced by
4670 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4672 38. Improved the line breaking for long SMTP error messages from ACLs.
4673 Previously, if there was no break point between 40 and 75 characters, Exim
4674 left the rest of the message alone. Two changes have been made: (a) I've
4675 reduced the minimum length to 35 characters; (b) if it can't find a break
4676 point between 35 and 75 characters, it looks ahead and uses the first one
4677 that it finds. This may give the occasional overlong line, but at least the
4678 remaining text gets split now.
4680 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4681 file descriptors might be low, and that setting 1000 would always raise it.
4682 It turns out that in some environments, the limit is already over 1000 and
4683 that lowering it causes trouble. So now Exim takes care not to decrease it.
4685 40. When delivering a message, the value of $return_path is set to $sender_
4686 address at the start of routing (routers may change the value). By an
4687 oversight, this default was not being set up when an address was tested by
4688 -bt or -bv, which affected the outcome if any router or filter referred to
4691 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4692 log only when "message" or "log_message" are set. However, if one of the
4693 conditions was an address verification, or a call to a nested ACL, the
4694 messages generated by the underlying test were being passed through. This
4695 no longer happens. The underlying message is available in $acl_verify_
4696 message for both "message" and "log_message" expansions, so it can be
4697 passed through if needed.
4699 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4700 new expansion $bh_ to give the encoded byte string without charset
4701 translation. Translation happens only if iconv() is available; HAVE_ICONV
4702 indicates this at build time. HEADERS_CHARSET gives the charset to
4703 translate to; headers_charset can change it in the configuration, and
4704 "headers charset" can change it in an individual filter file.
4706 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4707 that creates RFC 2047 encoded "words" labels them as that charset instead
4708 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4709 expansion operator; (ii) when Exim creates a From: line for a local
4710 message; (iii) when a header line is rewritten to include a "phrase" part.
4712 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4713 buggy, causing it to skip the first lines of messages whose message ID
4714 ended in 'D'. This would not have bitten before Exim release 4.14, because
4715 message IDs were unlikely to end in 'D' before then. The effect was to have
4716 incorrect size information for certain domains.
4718 45. #include "config.h" was missing at the start of the crypt16.c module. This
4719 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4722 46. If there was a timeout during a "random" callout check, Exim treated it as
4723 a failure of the random address, and carried on sending RSET and the real
4724 address. If the delay was just some slowness somewhere, the response to the
4725 original RCPT would be taken as a response to RSET and so on, causing
4726 mayhem of various kinds.
4728 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4729 when I implemented it. It didn't allow for the fact that some option values
4730 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4731 the right test for positive values.
4733 48. Domain names in DNS records are case-independent. Exim always looks them up
4734 in lower case. Some resolvers return domain names in exactly the case they
4735 appear in the zone file, that is, they may contain uppercase letters. Not
4736 all resolvers do this - some return always lower case. Exim was treating a
4737 change of case by a resolver as a change of domain, similar to a widening
4738 of a domain abbreviation. This triggered its re-routing code and so it was
4739 trying to route what was effectively the same domain again. This normally
4740 caused routing to fail (because the router wouldn't handle the domain
4741 twice). Now Exim checks for this case specially, and just changes the
4742 casing of the domain that it ultimately uses when it transmits the message
4745 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4748 50. If a filter generated a file delivery with a non-absolute name (possible if
4749 no home directory exists for the router), the forbid_file option was not
4752 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4753 the existing equality matching.
4755 52. Exim was using ints instead of ino_t variables in some places where it was
4756 dealing with inode numbers.
4758 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4759 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4760 called TMPDIR, and if it finds it is different, it changes its value.
4762 54. The smtp_printf() function is now made available to local_scan() so
4763 additional output lines can be written before returning. There is also an
4764 smtp_fflush() function to enable the detection of a dropped connection.
4765 The variables smtp_input and smtp_batched_input are exported to
4768 55. Changed the default runtime configuration: the message "Unknown user"
4769 has been removed from the ACL, and instead placed on the localuser router,
4770 using the cannot_route_message feature. This means that any verification
4771 failures that generate their own messages won't get overridden. Similarly,
4772 the "Unrouteable address" message that was in the ACL for unverifiable
4773 relay addresses has also been removed.
4775 56. Added hosts_avoid_esmtp to the smtp transport.
4777 57. The exicyclog script was not checking for the esoteric option
4778 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4779 will work only if exicyclog is run under the appropriate euid.
4781 58. Following a discussion on the list, the rules by which Exim recognises line
4782 endings on incoming messages have been changed. The -dropcr and drop_cr
4783 options are now no-ops, retained only for backwards compatibility. The
4784 following line terminators are recognized: LF CRLF CR. However, special
4785 processing applies to CR:
4787 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4788 nor a local message in the state where . is a terminator.
4790 (ii) If a bare CR is encountered in a header line, an extra space is added
4791 after the line terminator so as not to end the header. The reasoning
4792 behind this is that bare CRs in header lines are most likely either
4793 to be mistakes, or people trying to play silly games.
4795 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4796 was being incorrectly given as 18 bytes larger than it should have been.
4797 This is a VOB (very old bug).
4799 60. This may never have affected anything current, but just in case it has:
4800 When the local host is found other than at the start of a list of hosts,
4801 the local host, those with the same MX, and any that follow, are discarded.
4802 When the list in question was part of a longer list of hosts, the following
4803 hosts (not currently being processed) were also being discarded. This no
4804 longer happens. I'm not sure if this situation could ever has previously
4807 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4810 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4811 header. This is something that is recommended in a new Internet Draft, and
4812 is something that is documented as being done by Sendmail. There are two
4813 possible values. For messages generated by the autoreply transport, Exim
4816 Auto-Submitted: auto-replied
4818 whereas for all other generated messages (e.g. bounces) it adds
4820 Auto-Submitted: auto-generated
4822 63. The "personal" condition in filters now includes a test for the
4823 Auto-Submitted: header. If it contains the string "auto-" the message it
4824 not considered personal.
4826 64. Added rcpt_include_affixes as a generic transport option.
4828 65. Added queue_only_override (default true).
4830 66. Added the syslog_duplication option.
4832 67. If what should have been the first header line of a message consisted of
4833 a space followed by a colon, Exim was mis-interpreting it as a header line.
4834 It isn't of course - it is syntactically invalid and should therefore be
4835 treated as the start of the message body. The misbehaviour could have
4836 caused a number of strange effects, including loss of data in subsequent
4837 header lines, and spool format errors.
4839 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4840 client host had authenticated. This control can now be exercised by an ACL
4841 for more flexibility.
4843 69. By default, callouts do not happen when testing with -bh. There is now a
4844 variant, -bhc, which does actually run the callout code, including
4845 consulting and updating the callout cache.
4847 70. Added support for saslauthd authentication, courtesy of Alexander
4850 71. If statvfs() failed on the spool or log directories while checking their
4851 size for availability, Exim confusingly gave the error "space shortage".
4852 Furthermore, in debugging mode it crashed with a floating point exception.
4853 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4854 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4855 serious problem, Exim now writes to the main and panic logs when this
4856 happens, with details of the failure. It then refuses to accept the
4857 incoming message, giving the message "spool directory problem" or "log
4858 directory problem" with a 421 code for SMTP messages.
4860 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4861 0, 1, and 2 exist, because some OS complain for execs without them (see
4862 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4863 descriptors. However, the code omitted to check that the open succeeded,
4864 causing mysterious errors if for some reason the permissions on /dev/null
4865 got screwed. Now Exim writes a message to the main and panic logs, and
4866 bombs out if it can't open /dev/null.
4868 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4869 interact so that it is all more flexible. It is supposed to remain
4870 backwards compatible. Also added extra_local_interfaces.
4872 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4873 to bomb out with an assertion failure - to the client this appears as a
4874 connection drop. This problem occurs in the part of the code that was taken
4875 from the Samba project. Fortunately, the assertion is in a very simple
4876 function, so I have fixed this by reproducing the function inline in the
4877 one place where it is called, and arranging for authentication to fail
4878 instead of killing the process with assert().
4880 75. The SPA client code was not working when the server requested OEM rather
4881 than Unicode encoding.
4883 76. Added code to make require_files with a specific uid setting more usable in
4884 the case where statting the file as root fails - usually a non-root-mounted
4885 NFS file system. When this happens and the failure is EACCES, Exim now
4886 forks a subprocess and does the per-uid checking as the relevant uid.
4888 77. Added process_log_path.
4890 78. If log_file_path was not explicitly set, a setting of check_log_space or
4891 check_log_inodes was ignored.
4893 79. If a space check for the spool or log partitions fails, the incident is now
4894 logged. Of course, in the latter case the data may get lost...
4896 80. Added the %p formatting code to string_format() so that it can be used to
4897 print addresses in debug_print(). Adjusted all the address printing in the
4898 debugging in store.c to use %p rather than %d.
4900 81. There was a concern that a line of code in smtp_in.c could overflow a
4901 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4902 initially expressed, the concern was not well-founded, because trailing
4903 spaces are removed early. However, if the trailing spaces were followed by
4904 a NULL, they did not get removed, so the overflow was possible. Two fixes
4907 (a) I re-wrote the offending code in a cleaner fashion.
4908 (b) If an incoming SMTP command contains a NULL character, it is rejected
4911 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4912 runs initgroups(), so that if the Exim user is in any additional groups,
4913 they will be used during message reception.
4919 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob