1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.545 2008/02/14 13:49:35 fanf2 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
10 When building exim an external PCRE library is now needed -
11 PCRE is a system library on the majority of modern systems.
12 See entry on PCRE_LIBS in EDITME file.
14 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
15 conversation. Added nologin parameter to request.
16 Patch contributed by Kirill Miazine
18 TF/01 Do not log submission mode rewrites if they do not change the address.
20 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
22 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
23 log files in place. Contributed by Roberto Lima
25 NM/04 Bugzilla 667: close socket used by dovecot authenticator
27 TF/03 Bugzilla 615: When checking the local_parts router precondition
28 after a local_part_suffix or local_part_prefix option, Exim now
29 does not use the address's named list lookup cache, since this
30 contains cached lookups for the whole local part.
32 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
33 Robert Millan. Documentation is in experimental-spec.txt
35 TF/04 Bugzilla 668: Fix parallel build (make -j).
41 TK/01 Add preliminary DKIM support. Currently requires a forked version of
42 ALT-N's libdkim that I have put here:
43 http://duncanthrax.net/exim-experimental/
45 Note to Michael Haardt: I had to rename some vars in sieve.c. They
46 were called 'true' and it seems that C99 defines that as a reserved
47 keyword to be used with 'bool' variable types. That means you could
48 not include C99-style headers which use bools without triggering
49 build errors in sieve.c.
51 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
52 as mailq or other aliases. Changed the --help handling significantly
53 to do whats expected. exim_usage() emits usage/help information.
55 SC/01 Added the -bylocaldomain option to eximstats.
57 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
59 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
61 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
67 PH/01 Another patch from the Sieve maintainer.
69 PH/02 When an IPv6 address is converted to a string for single-key lookup
70 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
71 dots are used instead of colons so that keys in lsearch files need not
72 contain colons. This was done some time before quoting was made available
73 in lsearch files. However, iplsearch files do require colons in IPv6 keys
74 (notated using the quote facility) so as to distinguish them from IPv4
75 keys. This meant that lookups for IP addresses in host lists did not work
76 for iplsearch lookups.
78 This has been fixed by arranging for IPv6 addresses to be expressed with
79 colons if the lookup type is iplsearch. This is not incompatible, because
80 previously such lookups could never work.
82 The situation is now rather anomolous, since one *can* have colons in
83 ordinary lsearch keys. However, making the change in all cases is
84 incompatible and would probably break a number of configurations.
86 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
89 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
90 conversion specification without a maximum field width, thereby enabling
91 a rogue spamd server to cause a buffer overflow. While nobody in their
92 right mind would setup Exim to query an untrusted spamd server, an
93 attacker that gains access to a server running spamd could potentially
94 exploit this vulnerability to run arbitrary code as the Exim user.
96 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
97 $primary_hostname instead of what libspf2 thinks the hosts name is.
99 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
100 a directory entry by the name of the lookup key. Previously, if a
101 symlink pointed to a non-existing file or a file in a directory that
102 Exim lacked permissions to read, a lookup for a key matching that
103 symlink would fail. Now it is enough that a matching directory entry
104 exists, symlink or not. (Bugzilla 503.)
106 PH/03 The body_linecount and body_zerocount variables are now exported in the
109 PH/04 Added the $dnslist_matched variable.
111 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
112 This means they are set thereafter only if the connection becomes
115 PH/06 Added the client_condition to authenticators so that some can be skipped
116 by clients under certain conditions.
118 PH/07 The error message for a badly-placed control=no_multiline_responses left
119 "_responses" off the end of the name.
121 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
123 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
124 (without spaces) instead of just copying the configuration text.
126 PH/10 Added the /noupdate option to the ratelimit ACL condition.
128 PH/11 Added $max_received_linelength.
130 PH/12 Added +ignore_defer and +include_defer to host lists.
132 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
133 way in which PCRE > 7.0 is built.
135 PH/14 Implemented queue_only_load_latch.
137 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
138 MAIL command. The effect was to mangle the value on 64-bit systems.
140 PH/16 Another patch from the Sieve maintainer.
142 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
144 PH/18 If a system quota error occurred while trying to create the file for
145 a maildir delivery, the message "Mailbox is full" was not appended to the
146 bounce if the delivery eventually timed out. Change 4.67/27 below applied
147 only to a quota excession during the actual writing of the file.
149 PH/19 It seems that peer DN values may contain newlines (and other non-printing
150 characters?) which causes problems in log lines. The DN values are now
151 passed through string_printing() before being added to log lines.
153 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
154 and InterBase are left for another time.)
156 PH/21 Added message_body_newlines option.
158 PH/22 Guard against possible overflow in moan_check_errorcopy().
160 PH/23 POSIX allows open() to be a macro; guard against that.
162 PH/24 If the recipient of an error message contained an @ in the local part
163 (suitably quoted, of course), incorrect values were put in $domain and
164 $local_part during the evaluation of errors_copy.
170 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
171 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
174 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
175 issue a MAIL command.
177 PH/02 In an ACL statement such as
179 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
181 if a client was not listed at all, or was listed with a value other than
182 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
183 the condition was not true (as it should be), so access was not denied.
184 The bug was that the ! inversion was incorrectly passed on to the second
185 item. This has been fixed.
187 PH/03 Added additional dnslists conditions == and =& which are different from
188 = and & when the dns lookup returns more than one IP address.
190 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
191 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
193 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
194 FSYNC, which compiles an option called disable_fsync that allows for
195 bypassing fsync(). The documentation is heavily laced with warnings.
197 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
199 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
200 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
201 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
202 including adding "make clean"; (3) Added -fPIC when compiling the test
203 dynamically loaded module, to get rid of a warning.
205 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
206 message fails, move_frozen_messages = true and ignore_bounce_errors_after
207 = 0s. The bug is otherwise harmless.
209 PH/07 There was a bug in the dovecot authenticator such that the value of
210 $auth1 could be overwritten, and so not correctly preserved, after a
211 successful authentication. This usually meant that the value preserved by
212 the server_setid option was incorrect.
214 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
216 PH/09 Installed PCRE release 7.0.
218 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
219 run for batched SMTP input. It is now run at the start of every message
220 in the batch. While fixing this I discovered that the process information
221 (output by running exiwhat) was not always getting set for -bs and -bS
222 input. This is fixed, and it now also says "batched" for BSMTP.
224 PH/11 Added control=no_pipelining.
226 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
227 patch, slightly modified), and move the expansion of helo_data till after
228 the connection is made in the smtp transport (so it can use these
231 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
233 PH/14 Added log_selector = +pid.
235 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
237 PH/16 Add ${if forany and ${if forall.
239 PH/17 Added dsn_from option to vary the From: line in DSNs.
241 PH/18 Flush SMTP output before performing a callout, unless control =
242 no_callout_flush is set.
244 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
245 was true (the default) a successful delivery failed to delete the retry
246 item, thus causing premature timeout of the address. The bug is now
249 PH/20 Added hosts_avoid_pipelining to the smtp transport.
251 PH/21 Long custom messages for fakedefer and fakereject are now split up
252 into multiline reponses in the same way that messages for "deny" and
253 other ACL rejections are.
255 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
256 with slight modification.
258 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
259 draft, changing the syntax and factoring some duplicate code".
261 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
262 for deliveries of the second and subsequent messages over the same SMTP
265 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
266 ${reduce, with only minor "tidies".
268 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
270 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
271 expansion side effects.
273 PH/27 When a message times out after an over-quota error from an Exim-imposed
274 quota, the bounce message says "mailbox is full". This message was not
275 being given when it was a system quota that was exceeded. It now should
278 MH/03 Made $recipients available in local_scan(). local_scan() already has
279 better access to the recipient list through recipients_list[], but
280 $recipients can be useful in postmaster-provided expansion strings.
282 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
283 in the case of a MAIL command with additional options following the
284 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
285 were accidentally chopped off.
287 PH/29 SMTP synchronization checks are implemented when a command is read -
288 there is a check that no more input is waiting when there shouldn't be
289 any. However, for some commands, a delay in an ACL can mean that it is
290 some time before the response is written. In this time, more input might
291 arrive, invalidly. So now there are extra checks after an ACL has run for
292 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
293 pipelining has not been advertised.
295 PH/30 MH's patch to allow iscntrl() characters to be list separators.
297 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
298 returned in the SMTP response when smtp_return_error_details was false.
301 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
302 instead of the C I/O that was originally supplied, because problems were
305 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
306 Exim which did not show up earlier: it was assuming that a call to
307 SSL_CTX_set_info_callback() might give an error value. In fact, there is
308 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
309 was a macro that became an assignment, so it seemed to work. This has
310 changed to a proper function call with a void return, hence the compile
311 error. Exim's code has been fixed.
313 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
316 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
318 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
320 PH/37 If a message is not accepted after it has had an id assigned (e.g.
321 because it turns out to be too big or there is a timeout) there is no
322 "Completed" line in the log. When some messages of this type were
323 selected by exigrep, they were listed as "not completed". Others were
324 picked up by some special patterns. I have improved the selection
325 criteria to be more general.
327 PH/38 The host_find_failed option in the manualroute router can now be set
328 to "ignore", to completely ignore a host whose IP address cannot be
329 found. If all hosts are ignored, the behaviour is controlled by the new
330 host_all_ignored option.
332 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
333 homing or because of multiple MX records with /mx) generated more than
334 one IP address, and the following item turned out to be the local host,
335 all the secondary addresses of the first item were incorrectly removed
336 from the list, along with the local host and any following hosts (which
337 is what is supposed to happen).
339 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
340 whoever called Exim into the -H file. In the case of the daemon it was
341 behaving confusingly. When first started, it used values for whoever
342 started the daemon, but after a SIGHUP it used the Exim user (because it
343 calls itself on a restart). I have changed the code so that it now always
346 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
347 message are rejected with the same error (e.g. no authentication or bad
348 sender address), and a DATA command is nevertheless sent (as can happen
349 with PIPELINING or a stupid MUA), the error message that was given to the
350 RCPT commands is included in the rejection of the DATA command. This is
351 intended to be helpful for MUAs that show only the final error to their
354 PH/42 Another patch from the Sieve maintainer.
356 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
357 Eximstats - Fixed some broken HTML links and added missing column headers
359 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
360 columns for Rejects, Temp Rejects, Ham, and Spam rows.
362 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
364 PH/43 Yet another patch from the Sieve maintainer.
366 PH/44 I found a way to check for a TCP/IP connection going away before sending
367 the response to the final '.' that terminates a message, but only in the
368 case where the client has not sent further data following the '.'
369 (unfortunately, this is allowed). However, in many cases there won't be
370 any further data because there won't be any more messages to send. A call
371 to select() can be used: if it shows that the input is "ready", there is
372 either input waiting, or the socket has been closed. An attempt to read
373 the next input character can distinguish the two cases. Previously, Exim
374 would have sent an OK response which the client would never have see.
375 This could lead to message repetition. This fix should cure that, at
376 least in a lot of common cases.
378 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
379 advertised in response to EHLO.
385 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
386 fixed by 4.65/MH/01 (is this a record?) are fixed:
388 (i) An empty string was always treated as zero by the numeric comparison
389 operators. This behaviour has been restored.
391 (ii) It is documented that the numeric comparison operators always treat
392 their arguments as decimal numbers. This was broken in that numbers
393 starting with 0 were being interpreted as octal.
395 While fixing these problems I realized that there was another issue that
396 hadn't been noticed. Values of message_size_limit (both the global option
397 and the transport option) were treated as octal if they started with 0.
398 The documentation was vague. These values are now always treated as
399 decimal, and I will make that clear in the documentation.
405 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
406 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
409 MH/01 Don't check that the operands of numeric comparison operators are
410 integers when their expansion is in "skipping" mode (fixes bug
411 introduced by 4.64-PH/07).
413 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
414 child addresses, Exim now panics and dies. Previously, because the count
415 is held in a short int, deliveries were likely to be lost. As such a
416 large number of recipients for a single message is ridiculous
417 (performance will be very, very poor), I have chosen to impose a limit
418 rather than extend the field.
424 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
425 leftover -K file (the existence of which was triggered by #402).
426 While we were at it, introduced process PID as part of the -K
427 filename. This should rule out race conditions when creating
430 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
431 processing considerably. Previous code took too long for large mails,
432 triggering a timeout which in turn triggers #401.
434 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
435 in the DK code in transports.c. sendfile() is not really portable,
436 hence the _LINUX specificness.
438 TF/01 In the add_headers option to the mail command in an Exim filter,
439 there was a bug that Exim would claim a syntax error in any
440 header after the first one which had an odd number of characters
443 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
444 callout verification, Exim cached a "reject" for the entire domain. This
445 is correct for most verifications, but it is not correct for a recipient
446 verification with use_sender or use_postmaster set, because in that case
447 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
448 case of MAIL FROM:<> rejection from other early rejections (e.g.
449 rejection of HELO). When verifying a recipient using a non-null MAIL
450 address, the cache is ignored if it shows MAIL FROM:<> rejection.
451 Whatever the result of the callout, the value of the domain cache is
452 left unchanged (for any other kind of callout, getting as far as trying
453 RCPT means that the domain itself is ok).
455 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
458 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
459 manifest itself as EPIPE rather than ECONNECT. When tidying away a
460 session, the daemon ignores ECONNECT errors and logs others; it now
461 ignores EPIPE as well.
463 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
464 (quoted-printable decoding).
466 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
467 later the small subsequent patch to fix an introduced bug.
469 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
471 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
473 PH/08 An error is now given if message_size_limit is specified negative.
475 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
476 to be given (somewhat) arbitrary names.
478 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
481 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
482 miscellaneous code fixes
484 PH/10 Added the log_reject_target ACL modifier to specify where to log
487 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
488 hostname. This is wrong, because it relates to the incoming message (and
489 probably the interface on which it is arriving) and not to the outgoing
490 callout (which could be using a different interface). This has been
491 changed to use the value of the helo_data option from the smtp transport
492 instead - this is what is used when a message is actually being sent. If
493 there is no remote transport (possible with a router that sets up host
494 addresses), $smtp_active_hostname is used.
496 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
497 tweaks were necessary in order to get it to work (see also 21 below):
498 (a) The code assumed that strncpy() returns a negative number on buffer
499 overflow, which isn't the case. Replaced with Exim's string_format()
501 (b) There were several signed/unsigned issues. I just did the minimum
502 hacking in of casts. There is scope for a larger refactoring.
503 (c) The code used strcasecmp() which is not a standard C function.
504 Replaced with Exim's strcmpic() function.
505 (d) The code set only $1; it now sets $auth1 as well.
506 (e) A simple test gave the error "authentication client didn't specify
507 service in request". It would seem that Dovecot has changed its
508 interface. Fortunately there's a specification; I followed it and
509 changed what the client sends and it appears to be working now.
511 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
514 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
515 address A is aliased to B and C, where B exists and C does not. Without
516 -v the output is "A verified" because verification stops after a
517 successful redirection if more than one address is generated. However,
518 with -v the child addresses are also verified. Exim was outputting "A
519 failed to verify" and then showing the successful verification for C,
520 with its parentage. It now outputs "B failed to verify", showing B's
521 parentage before showing the successful verification of C.
523 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
524 look up a TXT record in a specific list after matching in a combined
527 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
528 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
529 they consult the DNS. I had assumed they would set it the way they
530 wanted; and indeed my experiments on Linux seem to show that in some
531 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
532 To be on the safe side, however, I have now made the interface to
533 host_find_byname() similar to host_find_bydns(), with an argument
534 containing the DNS resolver options. The host_find_byname() function now
535 sets these options at its start, just as host_find_bydns() does. The smtp
536 transport options dns_qualify_single and dns_search_parents are passed to
537 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
538 of host_find_byname() use the default settings of RES_DEFNAMES
539 (qualify_single) but not RES_DNSRCH (search_parents).
541 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
542 spool_read_header() do less string testing, by means of a preliminary
543 switch on the second character of optional "-foo" lines. (This is
544 overdue, caused by the large number of possibilities that now exist.
545 Originally there were few.) While I was there, I also converted the
546 str(n)cmp tests so they don't re-test the leading "-" and the first
547 character, in the hope this might squeeze out yet more improvement.
549 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
550 flag allowing group syntax was set by the header_syntax check but not
551 turned off, possible causing trouble later; (2) The flag was not being
552 set at all for the header_verify test, causing "group"-style headers to
553 be rejected. I have now set it in this case, and also caused header_
554 verify to ignore an empty address taken from a group. While doing this, I
555 came across some other cases where the code for allowing group syntax
556 while scanning a header line wasn't quite right (mostly, not resetting
557 the flag correctly in the right place). These bugs could have caused
558 trouble for malformed header lines. I hope it is now all correct.
560 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
561 with the "reply" argument non-NULL. The code, however (which originally
562 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
563 but it didn't always do it. This confused somebody who was copying the
564 code for some other use. I have removed all the tests.
566 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
567 feature that was used to support insecure browsers during the U.S. crypto
568 embargo. It requires special client support, and Exim is probably the
569 only MTA that supported it -- and would never use it because real RSA is
570 always available. This code has been removed, because it had the bad
571 effect of slowing Exim down by computing (never used) parameters for the
572 RSA_EXPORT functionality.
574 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
575 authenticator to fail if there's a tab character in the incoming data
576 (there should never be unless someone is messing about, as it's supposed
577 to be base64-encoded). Also added, on Timo's advice, the "secured" option
578 if the connection is using TLS or if the remote IP is the same as the
579 local IP, and the "valid-client-cert option" if a client certificate has
582 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
583 authenticators. This can be used for authorization after authentication
584 succeeds. (In the case of plaintext, it servers for both authentication
587 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
588 if any retry times were supplied.
590 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
591 connection, where there is no client IP address to check. In this
592 situation, the verify now always succeeds.
594 PH/25 Applied John Jetmore's -Mset patch.
596 PH/26 Added -bem to be like -Mset, but loading a message from a file.
598 PH/27 In a string expansion for a processed (not raw) header when multiple
599 headers of the same name were present, leading whitespace was being
600 removed from all of them, but trailing whitespace was being removed only
601 from the last one. Now trailing whitespace is removed from each header
602 before concatenation. Completely empty headers in a concatenation (as
605 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
606 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
608 PH/29 [Removed. This was a change that I later backed out, and forgot to
609 correct the ChangeLog entry (that I had efficiently created) before
610 committing the later change.]
612 PH/30 Exim was sometimes attempting to deliver messages that had suffered
613 address errors (4xx response to RCPT) over the same connection as other
614 messages routed to the same hosts. Such deliveries are always "forced",
615 so retry times are not inspected. This resulted in far too many retries
616 for the affected addresses. The effect occurred only when there were more
617 hosts than the hosts_max_try setting in the smtp transport when it had
618 the 4xx errors. Those hosts that it had tried were not added to the list
619 of hosts for which the message was waiting, so if all were tried, there
620 was no problem. Two fixes have been applied:
622 (i) If there are any address or message errors in an SMTP delivery, none
623 of the hosts (tried or untried) are now added to the list of hosts
624 for which the message is waiting, so the message should not be a
625 candidate for sending over the same connection that was used for a
626 successful delivery of some other message. This seems entirely
627 reasonable: after all the message is NOT "waiting for some host".
628 This is so "obvious" that I'm not sure why it wasn't done
629 previously. Hope I haven't missed anything, but it can't do any
630 harm, as the worst effect is to miss an optimization.
632 (ii) If, despite (i), such a delivery is accidentally attempted, the
633 routing retry time is respected, so at least it doesn't keep
634 hammering the server.
636 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
637 in ${readsocket because some servers need this prod.
639 PH/32 Added some extra debug output when updating a wait-xxx database.
641 PH/33 The hint "could be header name not terminated by colon", which has been
642 given for certain expansion errors for a long time, was not being given
643 for the ${if def:h_colon_omitted{... case.
645 PH/34 The spec says: "With one important exception, whenever a domain list is
646 being scanned, $domain contains the subject domain." There was at least
647 one case where this was not true.
649 PH/35 The error "getsockname() failed: connection reset by peer" was being
650 written to the panic log as well as the main log, but it isn't really
651 panic-worthy as it just means the connection died rather early on. I have
652 removed the panic log writing for the ECONNRESET error when getsockname()
655 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
656 runs only) independently of the message's sender address. This meant
657 that, if the 4xx error was in fact related to the sender, a different
658 message to the same recipient with a different sender could confuse
659 things. In particualar, this can happen when sending to a greylisting
660 server, but other circumstances could also provoke similar problems.
661 I have changed the default so that the retry time for these errors is now
662 based a combination of the sender and recipient addresses. This change
663 can be overridden by setting address_retry_include_sender=false in the
666 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
667 remote server are returned as part of bounce messages. This was not
668 happening for LMTP over a pipe (the lmtp transport), but now it is the
669 same for both kinds of LMTP.
671 PH/38 Despite being documented as not happening, Exim was rewriting addresses
672 in header lines that were in fact CNAMEs. This is no longer the case.
674 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
675 and queue runs started by the daemon processed all messages. This has
676 been fixed so that -R and -S can now usefully be given with -q<time>.
678 PH/40 Import PCRE release 6.7 (fixes some bugs).
680 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
682 PH/42 Give an error if -q is specified more than once.
684 PH/43 Renamed the variables $interface_address and $interface_port as
685 $received_ip_address and $received_port, to make it clear that these
686 values apply to message reception, and not to the outgoing interface when
687 a message is delivered. (The old names remain recognized, of course.)
689 PH/44 There was no timeout on the connect() call when using a Unix domain
690 socket in the ${readsocket expansion. There now is.
692 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
693 be meaningful with "accept".
695 SC/01 Eximstats V1.43
696 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
698 SC/02 Eximstats V1.44
699 Use a glob alias rather than an array ref in the generated
700 parser. This improves both readability and performance.
702 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
703 Collect SpamAssassin and rejection statistics.
704 Don't display local sender or destination tables unless
705 there is data to show.
706 Added average volumes into the top table text output.
708 SC/04 Eximstats V1.46
709 Collect data on the number of addresses (recipients)
710 as well as the number of messages.
712 SC/05 Eximstats V1.47
713 Added 'Message too big' to the list of mail rejection
714 reasons (thanks to Marco Gaiarin).
716 SC/06 Eximstats V1.48
717 Mainlog lines which have GMT offsets and are too short to
718 have a flag are now skipped.
720 SC/07 Eximstats V1.49 (Alain Williams)
721 Added the -emptyok flag.
723 SC/08 Eximstats V1.50
724 Fixes for obtaining the IP address from reject messages.
726 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
727 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
728 whitesspace changes from 4.64-PH/27
730 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
733 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
734 are found, allow negative numbers in numeric criteria)
736 JJ/06 exipick.20061117.2, added new $message_body_missing variable
738 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
739 to match changes made in 4.64-PH/43
741 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
743 PH/47 Put in an explicit test for a DNS lookup of an address record where the
744 "domain" is actually an IP address, and force a failure. This locks out
745 those revolvers/nameservers that support "A-for-A" lookups, in
746 contravention of the specifications.
748 PH/48 When a host name was looked up from an IP address, and the subsequent
749 forward lookup of the name timed out, the host name was left in
750 $sender_host_name, contrary to the specification.
752 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
753 restricted to single-key lookups, Exim was not diagnosing an error if
754 * or *@ was used with a query-style lookup.
756 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
758 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
759 long ago, but noone interested enough thought of it. Let's just say that
760 the "1.1" means that there are some new functions that weren't there at
761 some point in the past.
763 PH/51 Error processing for expansion failure of helo_data from an smtp
764 transport during callout processing was broken.
766 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
767 tested/used via the -bh/-bhc/-bs options.
769 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
770 bug, fixed in subsequent PCRE releases).
772 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
773 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
775 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
781 SC/01 Use a glob alias rather than an array ref in eximstats generated
782 parser. This improves both readability and performance.
784 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
785 Don't display local sender or destination tables in eximstats unless
786 there is data to show.
787 Added average volumes into the eximstats top table text output.
789 SC/03 Collect data on the number of addresses (recipients) as well
790 as the number of messages in eximstats.
792 TF/01 Correct an error in the documentation for the redirect router. Exim
793 does (usually) call initgroups() when daemonizing.
795 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
796 with consistent privilege compared to when running as a daemon.
798 TF/03 Note in the spec that $authenticated_id is not set for local
799 submissions from trusted users.
801 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
802 Thanks to Dean Brooks <dean@iglou.com> for the patch.
804 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
805 by adding some example configuration directives to the default
806 configuration file. A little bit of work is required to uncomment the
807 directives and define how usernames and passwords are checked, but
808 there is now a framework to start from.
810 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
811 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
812 without this. I don't know how relevant this is to other LDAP libraries.
814 PH/02 Add the verb name to the "unknown ACL verb" error.
816 PH/03 Magnus Holmgren's patch for filter_prepend_home.
818 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
820 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
821 directory not expanded when it should be if an expanded home directory
822 was set for the address (which is overridden by the transport).
824 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
827 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
828 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
829 because it is too late at that time, and has no effect.
831 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
832 security issue with \' (bugzilla #107). I could not use the
833 PQescapeStringConn() function, because it needs a PGconn value as one of
836 PH/08 When testing addresses using -bt, indicate those final addresses that
837 are duplicates that would not cause an additional delivery. At least one
838 person was confused, thinking that -bt output corresponded to deliveries.
839 (Suppressing duplicates isn't a good idea as you lose the information
840 about possibly different redirections that led to the duplicates.)
842 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
843 systems where poll() doesn't work, in particular OS X.
845 PH/10 Added more information to debugging output for retry time not reached.
847 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
848 operations in malware.c.
850 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
853 PH/13 If write_rejectlog was set false when logging was sent to syslog with
854 syslog_duplication set false, log lines that would normally be written
855 both the the main log and to the reject log were not written to syslog at
858 PH/14 In the default configuration, change the use of "message" in ACL warn
859 statements to "add_header".
861 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
862 not followed by a command (e.g. "seen endif").
864 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
865 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
868 PH/17 Added extra conditions to the default value of delay_warning_condition
872 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
873 { match{$h_precedence:}{(?i)bulk|list|junk} } \
874 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
877 The Auto-Submitted: and various List- headers are standardised, whereas I
878 don't think Precedence: ever was.
880 PH/18 Refactored debugging code in route_finduser() to show more information,
881 in particular, the error code if getpwnam() issues one.
883 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
884 This is apparently needed in addition to the PH/07 change above to avoid
885 any possible encoding problems.
887 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
888 but not after initializing Perl.
890 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
891 output them only if debugging. By default they are written stderr,
892 apparently, which is not desirable.
894 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
897 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
900 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
902 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
903 authenticated or an ident call has been made. Suppress the default
904 values for $authenticated_id and $authenticated_sender (but permit -oMai
905 and -oMas) when testing with -bh.
907 PH/24 Re-jigged the order of the tests in the default configuration so that the
908 tests for valid domains and recipients precede the DNS black list and CSA
909 tests, on the grounds that those ones are more expensive.
911 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
912 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
913 This bug exists in every version of Exim that I still have, right back to
916 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
917 However, an attempt to turn on case-sensitivity in a regex key by
918 including (?-i) didn't work because the subject string was already
919 lowercased, and the effects were non-intuitive. It turns out that a
920 one-line patch can be used to allow (?-i) to work as expected.
926 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
927 other effects) broke the use of negated acl sub-conditions.
929 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
932 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
933 "Deny" causes Exim to reject the incoming connection with a 554 error.
934 Unfortunately, if there is a major crisis, such as a disk failure,
935 tcp-wrappers gives "deny", whereas what one would like would be some
936 kind of temporary error. A kludge has been added to help with this.
937 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
938 554 error is used if errno is still zero or contains ENOENT (which occurs
939 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
942 PH/03 Add -lutil to the default FreeBSD LIBS setting.
944 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
945 errors. Otherwise a message that provokes a temporary error (when other
946 messages do not) can cause a whole host to time out.
948 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
949 addresses were routed directly to files or pipes from a redirect router.
950 File deliveries just didn't batch; pipe deliveries might have suffered
953 PH/06 A failure to get a lock for a hints database would erroneously always say
954 "Failed to get write lock", even when it was really a read lock.
956 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
957 of 0600. This has been changed to use the value of the lockfile_mode
958 option (which defaults to 0600).
960 PH/08 Applied small patch from the Sieve maintainer.
962 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
963 folder from quota calculations, a direct delivery into this folder messed
964 up the contents of the maildirsize file. This was because the regex was
965 used only to exclude .Trash (or whatever) when the size of the mailbox
966 was calculated. There was no check that a delivery was happening into an
967 excluded directory. This bug has been fixed by ignoring all quota
968 processing for deliveries into excluded directories.
970 PH/10 Added the maildirfolder_create_regex option to appendfile.
976 PH/01 The code for finding all the local interface addresses on a FreeBSD
977 system running IPv6 was broken. This may well have applied to all BSD
978 systems, as well as to others that have similar system calls. The broken
979 code found IPv4 interfaces correctly, but gave incorrect values for the
980 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
981 that it would not match correctly against @[] and not recognize the IPv6
984 PH/02 The ipliteral router was not recognizing addresses of the form user@
985 [ipv6:....] because it didn't know about the "ipv6:" prefix.
987 PH/03 Added disable_ipv6.
989 PH/04 Changed $reply_address to use the raw form of the headers instead of the
990 decoded form, because it is most often used to construct To: headers
991 lines in autoreplies, and the decoded form may well be syntactically
992 invalid. However, $reply_address has leading white space removed, and all
993 newlines turned into spaces so that the autoreply transport does not
996 PH/05 If group was specified without a user on a router, and no group or user
997 was specified on a transport, the group from the router was ignored.
999 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1000 for visible compile-time settings that can be used to change these
1001 numbers, for those that want even more. Backwards compatibility with old
1002 spool files has been maintained. However, going back to a previous Exim
1003 release will lost any variables that are in spool files.
1005 PH/07 Two small changes when running in the test harness: increase delay when
1006 passing a TCP/IP connection to a new process, in case the original
1007 process has to generate a bounce, and remove special handling of
1008 127.0.0.2 (sic), which is no longer necessary.
1010 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1011 be the same on different OS.
1013 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1016 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1017 whether --show-vars was specified or not
1019 JJ/02 exipick: Added support for new ACL variable spool format introduced
1022 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1023 syntactically invalid From: or Reply-to: line, and a filter used this to
1024 generate an autoreply, and therefore failed to obtain an address for the
1025 autoreply, Exim could try to deliver to a non-existent relative file
1026 name, causing unrelated and misleading errors. What now happens is that
1027 it logs this as a hard delivery error, but does not attempt to create a
1030 PH/11 The exinext utility has a -C option for testing purposes, but although
1031 the given file was scanned by exinext itself; it wasn't being passed on
1032 when Exim was called.
1034 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1035 an end-of-file indication when reading a command response.
1037 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1038 compiled. In many other places in Exim, IPv6 addresses are always
1039 recognized, so I have changed this. It also means that IPv4 domain
1040 literals of the form [IPV4:n.n.n.n] are now always recognized.
1042 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1043 used if the router is not running as root, for example, when verifying at
1044 ACL time, or when using -bh. The debugging output from this situation was
1045 non-existent - all you got was a failure to exec. I have made two
1048 (a) Failures to set uid/gid, the current directory, or a process leader
1049 in a subprocess such as that created by queryprogram now generate
1050 suitable debugging ouput when -d is set.
1052 (b) The queryprogram router detects when it is not running as root,
1053 outputs suitable debugging information if -d is set, and then runs
1054 the subprocess without attempting to change uid/gid.
1056 PH/15 Minor change to Makefile for building test_host (undocumented testing
1059 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1060 additional section of a DNS packet that returns MX or SRV records.
1061 Instead, it always explicitly searches for A/AAAA records. This avoids
1062 major problems that occur when a DNS server includes only records of one
1063 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1064 fixed another bug: if SRV records were looked up and the corresponding
1065 address records were *not* found in the additional section, the port
1066 values from the SRV records were lost.
1068 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1069 using the correct key (the original address) when searching the retry
1070 rules in order to find which one to use for generating the retry hint.
1072 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1073 adding the default one. Similarly, if it contains a Reply-To: header, the
1074 errors_reply_to option, if set, is not used.
1076 PH/19 When calculating a retry time, Exim used to measure the "time since
1077 failure" by looking at the "first failed" field in the retry record. Now
1078 it does not use this if it is later than than the arrival time of the
1079 message. Instead it uses the arrival time. This makes for better
1080 behaviour in cases where some deliveries succeed, thus re-setting the
1081 "first failed" field. An example is a quota failure for a huge message
1082 when small messages continue to be delivered. Without this change, the
1083 "time since failure" will always be short, possible causing more frequent
1084 delivery attempts for the huge message than are intended.
1085 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1087 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1088 $1, $2, $3) because the numerical variables can be reset during some
1089 expansion items (e.g. "match"), thereby losing the authentication data.
1091 PH/21 Make -bV show the size of off_t variables so that the test suite can
1092 decide whether to run tests for quotas > 2G.
1094 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1095 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1096 filecount value is greater than 2G or if a quota value is greater than 2G
1097 on a system where the size of off_t is not greater than 4, a panic error
1100 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1101 never match. The debug and -bh output now contains an explicit error
1102 message indicating a malformed IPv4 address or mask.
1104 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1105 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1106 PH/23 above applies.
1108 PH/25 Do not write to syslog when running in the test harness. The only
1109 occasion when this arises is a failure to open the main or panic logs
1110 (for which there is an explicit test).
1112 PH/26 Added the /no_tell option to "control=freeze".
1114 PH/27 If a host name lookup failed very early in a connection, for example, if
1115 the IP address matched host_lookup and the reverse lookup yielded a name
1116 that did not have a forward lookup, an error message of the form "no IP
1117 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1118 could be logged. Now it outputs the IP address instead of "NULL".
1120 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1121 allows the sender and the authenticated sender to be set when
1122 submitting a message from within Exim. Since child_open_exim() is
1123 documented for local_scan(), the new function should be too.
1125 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1126 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1127 results in an empty string is now treated as unset.
1129 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1131 PH/31 Added sender_verify_fail logging option.
1133 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1134 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1135 tidied the source and removed it altogether.
1137 PH/33 When a queue run was abandoned because the load average was too high, a
1138 log line was always written; now it is written only if the queue_run log
1139 selector is set. In addition, the log line for abandonment now contains
1140 information about the queue run such as the pid. This is always present
1141 in "start" and "stop" lines but was omitted from the "abandon" line.
1143 PH/34 Omit spaces between a header name and the colon in the error message that
1144 is given when verify = headers_syntax fails (if there are lots of them,
1145 the message gets confusing).
1147 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1148 names, as there are now some PTR records that contain slashes. This check
1149 is only to protect against broken name servers that fall over on strange
1150 characters, so the fact that it applies to all lookups doesn't matter.
1152 PH/36 Now that the new test suite is complete, we can remove some of the
1153 special code in Exim that was needed for the old test suite. For example,
1154 sorting DNS records because real resolvers return them in an arbitrary
1155 order. The new test suite's fake resolver always returns records in the
1158 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1159 bounces) except when queue_only is set, to avoid logging races between
1160 the different processes.
1162 PH/38 Panic-die if .include specifies a non-absolute path.
1164 PH/39 A tweak to the "H" retry rule from its user.
1166 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1167 a label. They prevented compilation on older perls.
1169 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1170 a warning to be raised on newish perls.
1172 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1173 on queue. Changes to match documented behaviour of showing count of
1174 messages matching specified criteria.
1176 PH/40 Changed the default ident timeout from 30s to 5s.
1178 PH/41 Added support for the use of login_cap features, on those BSD systems
1179 that have them, for controlling the resources used by pipe deliveries.
1181 PH/42 The content-scanning code uses fopen() to create files in which to put
1182 message data. Previously it was not paying any attention to the mode of
1183 the files. Exim runs with umask(0) because the rest of the code creates
1184 files with open(), and sets the required mode explicitly. Thus, these
1185 files were ending up world-writeable. This was not a big issue, because,
1186 being within the spool directory, they were not world-accessible. I have
1187 created a function called modefopen, which takes an additional mode
1188 argument. It sets umask(777), creates the file, chmods it to the required
1189 mode, then resets the umask. All the relevant calls to fopen() in the
1190 content scanning code have been changed to use this function.
1192 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1193 to 24 hours. This avoids potential overflow problems when processing G
1194 and H retry rules. I suspect nobody ever tinkers with this value.
1196 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1198 PH/45 When the plaintext authenticator is running as a client, the server's
1199 challenges are checked to ensure they are valid base64 strings. By
1200 default, the authentication attempt is cancelled if an invalid string is
1201 received. Setting client_ignore_invalid_base64 true ignores these errors.
1202 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1203 they are received. Thus, the responses can be made to depend on the
1204 challenges. If an invalid string is ignored, an empty string is placed in
1207 PH/46 Messages that are created by the autoreply transport now contains a
1208 References: header, in accordance with RFCs 2822 and 3834.
1210 PH/47 Added authenticated_sender_force to the smtp transport.
1212 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1214 PH/49 Installed latest patch from the Sieve maintainer.
1216 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1217 was also set, the appendfile transport was unnecessarily scanning a
1218 directory of message files (e.g. for maildir delivery) to find the count
1219 of files (along with the size), even though it did not need this
1220 information. It now does the scan only if it needs to find either the
1221 size of the count of files.
1223 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1225 PH/52 Two bugs concerned with error handling when the smtp transport is
1228 (i) Exim was not creating retry information for temporary errors given
1229 for individual recipients after the DATA command when the smtp transport
1230 was used in LMTP mode. This meant that they could be retried too
1231 frequently, and not timed out correctly.
1233 (ii) Exim was setting the flag that allows error details to be returned
1234 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1235 recipients that were returned after the DATA command.
1237 PH/53 This is related to PH/52, but is more general: for any failing address,
1238 when detailed error information was permitted to be returned to the
1239 sender, but the error was temporary, then after the final timeout, only
1240 "retry timeout exceeded" was returned. Now it returns the full error as
1241 well as "retry timeout exceeded".
1243 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1244 do this, and (what is worse) MTAs that accept it.
1246 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1247 will now be deprecated.
1249 PH/56 New os.c-cygwin from the Cygwin maintainer.
1251 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1252 formats (previously only available in exim formats via -bpr, -bpru,
1253 and -bpra. Now also available in native and exiqgrep formats)
1255 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1256 with very large, slow to parse queues
1258 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1260 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1262 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1263 responses to authentication challenges, though it was showing the
1264 challenges; (ii) I've removed the CR characters from the debug output for
1267 PH/58 Allow for the insertion of a newline as well as a space when a string
1268 is turned into more than one encoded-word during RFC 2047 encoding. The
1269 Sieve code now uses this.
1271 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1272 data_4xx, lost_connection, tls_required.
1274 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1275 message was being sent as an SMTP response.
1277 PH/61 Add -l and -k options to exicyclog.
1279 PH/62 When verifying, if an address was redirected to one new address, so that
1280 verification continued, and the new address failed or deferred after
1281 having set something in $address_data, the value of $address_data was not
1282 passed back to the ACL. This was different to the case when no
1283 redirection occurred. The value is now passed back in both cases.
1285 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1286 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1287 use login_cap.h, so on its own it isn't the distinguishing feature. The
1288 new name refers directly to the setclassresources() function.
1290 PH/65 Added configuration files for NetBSD3.
1292 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1294 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1295 is preferred over IPv4.
1297 PH/68 The bounce_return_message and bounce_return_body options were not being
1298 honoured for bounces generated during the reception of non-SMTP messages.
1299 In particular, this applied to messages rejected by the ACL. This bug has
1300 been fixed. However, if bounce_return_message is true and bounce_return_
1301 body is false, the headers that are returned for a non-SMTP message
1302 include only those that have been read before the error was detected.
1303 (In the case of an ACL rejection, they have all been read.)
1305 PH/69 The HTML version of the specification is now built in a directory called
1306 spec_html instead of spec.html, because the latter looks like a path with
1307 a MIME-type, and this confuses some software.
1309 PH/70 Catch two compiler warnings in sieve.c.
1311 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1312 function verify_get_ident() calls ip_connect() to connect a socket, but
1313 if the "connect()" function timed out, ip_connect() used to close the
1314 socket. However, verify_get_ident() also closes the socket later, and in
1315 between Exim writes to the log, which may get opened at this point. When
1316 the socket was closed in ip_connect(), the log could get the same file
1317 descriptor number as the socket. This naturally causes chaos. The fix is
1318 not to close the socket in ip_connect(); the socket should be closed by
1319 the function that creates it. There was only one place in the code where
1320 this was missing, in the iplookup router, which I don't think anybody now
1321 uses, but I've fixed it anyway.
1323 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1324 well as to direct DNS lookups. Otherwise the handling of names in host
1325 lists is inconsistent and therefore confusing.
1331 PH/01 Two changes to the default runtime configuration:
1333 (1) Move the checks for relay_from_hosts and authenticated clients from
1334 after to before the (commented out) DNS black list checks.
1336 (2) Add control=submission to the relay_from_hosts and authenticated
1337 clients checks, on the grounds that messages accepted by these
1338 statements are most likely to be submissions.
1340 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1342 (1) Generate an error if the third argument for the ${prvs expansion is
1345 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1348 (3) Reset the variables that are obtained from the first argument of
1349 ${prvscheck and used in the second argument before leaving the code,
1350 because their memory is reclaimed, so using them afterwards may do
1353 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1354 one (it's much easier than Tom thought :-).
1356 (5) Because of (4), we can now allow for the use of $prvscheck_result
1357 inside the third argument.
1359 PH/03 For some reason, the default setting of PATH when running a command from
1360 a pipe transport was just "/usr/bin". I have changed it to
1363 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1364 anything to be listed in the output from -bV.
1366 PH/05 When a filter generated an autoreply, the entire To: header line was
1367 quoted in the delivery log line, like this:
1369 => >A.N.Other <ano@some.domain> <original@ddress> ...
1371 This has been changed so that it extracts the operative address. There
1372 may be more than one such address. If so, they are comma-separated, like
1375 => >ano@some.domain,ona@other.domain <original@ddress> ...
1377 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1378 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1379 not being looked up in the rDNS to get a host name, Exim was showing the
1380 IP address twice in Received: lines, even though the IP addresses were
1381 identical. For example:
1383 Received: from [1.2.3.4] (helo=[1.2.3.4])
1385 However, if the real host name was known, it was omitting the HELO data
1386 if it matched the actual IP address. This has been tidied up so that it
1387 doesn't show the same IP address twice.
1389 PH/07 When both +timestamp and +memory debugging was on, the value given by
1390 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1391 called by the debug printing, thereby overwriting the timestamp buffer.
1392 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1394 PH/08 When the original message was included in an autoreply transport, it
1395 always said "this is a copy of the message, including all the headers",
1396 even if body_only or headers_only was set. It now gives an appropriate
1399 PH/09 Applied a patch from the Sieve maintainer which:
1401 o fixes some comments
1402 o adds the (disabled) notify extension core
1403 o adds some debug output for the result of if/elsif tests
1404 o points to the current vacation draft in the documentation
1405 and documents the missing references header update
1409 o fixes a bug in processing the envelope test (when testing
1410 multiple envelope elements, the last element determinted the
1413 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1414 Electronic Mail") by including:
1416 Auto-submitted: auto-generated
1418 in the messages that it generates (bounce messages and others, such as
1419 warnings). In the case of bounce messages for non-SMTP mesages, there was
1420 also a typo: it was using "Auto_submitted" (underscore instead of
1421 hyphen). Since every message generated by Exim is necessarily in response
1422 to another message, thes have all been changed to:
1424 Auto-Submitted: auto-replied
1426 in accordance with these statements in the RFC:
1428 The auto-replied keyword:
1430 - SHOULD be used on messages sent in direct response to another
1431 message by an automatic process,
1433 - MUST NOT be used on manually-generated messages,
1435 - MAY be used on Delivery Status Notifications (DSNs) and Message
1436 Disposition Notifications (MDNs),
1438 - MUST NOT be used on messages generated by automatic or periodic
1439 processes, except for messages which are automatic responses to
1442 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1443 to the default Received: header definition.
1445 PH/12 Added log selector acl_warn_skipped (default on).
1447 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1448 variables because (a) they are in the wrong storage pool and (b) even if
1449 they were copied, it wouldn't work properly because of the caching.
1451 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1452 checking when decoding. Apparently there are clients that generate
1453 overlong encoded strings. Why am I not surprised?
1455 PH/15 If the first argument of "${if match_address" was not empty, but did not
1456 contain an "@" character, Exim crashed. Now it writes a panic log message
1457 and treats the condition as false.
1459 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1461 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1462 extension "envelope-auth". The code is finished and in agreement with
1463 other implementations, but there is no documentation so far and in fact,
1464 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1465 not changing the active code.
1467 Print executed "if" and "elsif" statements when debugging is used. This
1468 helps a great deal to understand what a filter does.
1470 Document more things not specified clearly in RFC3028. I had all this
1471 sorted out, when out of a sudden new issues came to my mind. Oops."
1473 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1476 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1477 canonical form (as documented). However, after a host name lookup from
1478 the IP address, check_host() was doing a simple string comparison with
1479 addresses acquired from the DNS when checking that the found name did
1480 have the original IP as one of its addresses. Since any found IPv6
1481 addresses are likely to be in abbreviated form, the comparison could
1482 fail. Luckily, there already exists a function for doing the comparison
1483 by converting both addresses to binary, so now that is used instead of
1484 the text comparison.
1486 PH/20 There was another similar case to PH/19, when a complete host name was
1487 given in a host list; looking up its IP address could give an abbreviated
1488 form, whereas the current host's name might or might not be abbreviated.
1489 The same fix has been applied.
1495 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1496 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1499 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1500 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1502 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1504 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1505 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1506 Also, for "Auto-Submitted", treat anything other than "no" as
1507 non-personal, in accordance with RFC 3834. (Previously it treated
1508 anything starting "auto-" as non-personal.)
1510 TF/01 The control=submission/name=... option had a problem with syntax
1511 errors if the name included a slash character. The /name= option
1512 now slurps the rest of the string, so it can include any characters
1513 but it must come last in the list of options (after /sender_retain
1516 PH/05 Some modifications to the interface to the fake nameserver for the new
1524 TK/01 Added the "success_on_redirect" address verification option. See
1525 NewStuff for rationale and an example.
1527 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1529 PH/02 Patch to exigrep to allow it to work on syslog lines.
1531 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1532 fread() to skip over the body file's header line, because in Cygwin the
1533 header line is locked and is inaccessible.
1535 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1536 co-exist for some time) to make it clear that it is the Exim ID that is
1537 referenced, not the Message-ID: header line.
1539 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1540 string_format() function, because snprintf() does not exist on all
1543 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1544 ${stat: expansion item.
1546 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1547 protocol synchronization error", to keep the pedants happy.
1549 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1550 well as for IRIX systems, when gcc is being used. See the host.c source
1553 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1555 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1558 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1559 transport and to the smtp transport in LMTP mode.
1561 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1563 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1564 run a filter in a subprocess. This could lead to confusion in subsequent
1565 lookups in the parent process. There should also be a search_tidyup() at
1566 the end of the subprocess.
1568 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1569 only if the host matched helo_try_verify_hosts, which caused the
1570 verification to occur when the EHLO/HELO command was issued. The ACL just
1571 tested the remembered result. Now, if a previous verification attempt has
1572 not happened, "verify = helo" does it there and then.
1574 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1576 TK/03 Fix log output including CR from clamd.
1578 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1579 exist provoked a memory error which could cause a segfault.
1581 PH/15 Installed PCRE 6.2
1583 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1585 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1586 of the problem. Specifically, suggested +O2 rather than +O1 for the
1589 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1591 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1592 record, the port was not correctly logged when the outgoing_port log
1593 selector was set (it logged the transort's default port).
1595 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1596 fallback_hosts, and "hosts" in the smtp transport.
1598 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1599 host errors such as "Connection refused".
1601 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1602 authentication with radiusclient 0.4.9:
1604 - Error returned from rc_read_config was caught wrongly
1605 - Username/password not passed on to radius server due to wrong length.
1607 The presumption is that some radiusclient API changes for 4.51/PH/17
1608 were not taken care of correctly. The code is still untested by me (my
1609 Linux distribution still has 0.3.2 of radiusclient), but it was
1610 contributed by a Radius user.
1612 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1613 expanding the "port" option of the smtp transport.
1615 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1616 while reading a MIME header. Thanks to Tom Hughes for a patch.
1618 PH/24 Include config.h inside local_scan.h so that configuration settings are
1621 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1622 that in an ACL for RCPT (for example), you can examine exactly what was
1625 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1626 commands, but it was not correctly comparing the address with the actual
1627 client host address. Thus, it would show the EHLO address in Received:
1628 header lines when this was not necessary.
1630 PH/27 Added the % operator to ${eval:}.
1632 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1633 it should be ignoring failures (because with -C, for example, it has lost
1634 privilege). It wasn't ignoring creation failures other than "already
1637 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1640 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1641 because an input line was too long, either on its own, or by virtue of
1642 too many continuations, the temporary file was not being removed, and the
1643 return code was incorrect.
1645 PH/31 Missing "BOOL" in function definition in filtertest.c.
1647 PH/32 Applied Sieve patches from the maintainer.
1649 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1651 PH/33 Added "verify = not_blind".
1653 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1654 Local/Makefile (with some defaults set). These are used in built scripts
1655 such as exicyclog, but they have never been used in the exim_install
1656 script (though there are many overriding facilities there). I have
1657 arranged that the exim_install script now takes note of these two
1660 PH/35 Installed configuration files for Dragonfly.
1662 PH/36 When a locally submitted message by a trusted user did not contain a
1663 From: header, and the sender address was obtained from -f or from an SMTP
1664 MAIL command, and the trusted user did not use -F to supply a sender
1665 name, $originator_name was incorrectly used when constructing a From:
1666 header. Furthermore, $originator_name was used for submission mode
1667 messages from external hosts without From: headers in a similar way,
1668 which is clearly wrong.
1670 PH/37 Added control=suppress_local_fixups.
1672 PH/38 When log_selector = +received_sender was set, and the addition of the
1673 sender made the log line's construction buffer exactly full, or one byte
1674 less than full, an overflow happened when the terminating "\n" was
1677 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1678 when the result of a list match is failure because a DNS lookup failed.
1680 PH/40 RM_COMMAND is now used in the building process.
1682 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1683 the "build-* directories that it finds.
1685 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1686 address in a domain literal was a prefix of an interface address.
1688 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1689 when verifying a sender address, unless rewrite_headers is false.
1691 PH/44 Wrote a long comment about why errors_to addresses are verified as
1692 recipients, not senders.
1694 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1695 the ratelimit ACL was added.
1697 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1699 PH/46 Added extra information about PostgreSQL errors to the error string.
1701 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1702 avoiding the need to install special zones in a real server. This is
1703 backwards compatible; if it can't find the fake resolver, it drops back.
1704 Thus, both old and new test suites can be run.
1706 TF/02 Added util/ratelimit.pl
1708 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1709 clock is set back in time.
1711 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1712 Candler <B.Candler@pobox.com>.
1714 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1715 OK for addresses that are the result of redirections.
1717 PH/48 A number of further additions for the benefit of the new test suite,
1718 including a fake gethostbyname() that interfaces to the fake DNS resolver
1721 TF/06 The fix for widen_domains has also been applied to qualify_single and
1722 search_parents which are the other dnslookup options that can cause
1725 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1728 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1730 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1731 service" if it encountered an MX record with an empty target hostname.
1732 The message is now "an MX or SRV record indicated no SMTP service".
1734 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1735 if the DNS of the sending site is misconfigured. This is quite a
1736 common situation. This change restores the behaviour of treating a
1737 helo verification defer as a failure.
1739 PH/51 If self=fail was set on a router, the bounce message did not include the
1740 actual error message.
1746 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1748 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1749 command itself ended in error, the underlying message about the transport
1750 filter timeout was being overwritten with the pipe command error. Now the
1751 underlying error message should be appended to the second error message.
1753 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1755 PH/02 Reduce the amount of output that "make" produces by default. Full output
1756 can still be requested.
1758 PH/03 The warning log line about a condition test deferring for a "warn" verb
1759 was being output only once per connection, rather than after each
1760 occurrence (because it was using the same function as for successful
1761 "warn" verbs). This seems wrong, so I have changed it.
1763 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1764 it should not have, which might have caused a crash in the right
1765 circumstances, but probably never did.
1767 PH/04 Installed a modified version of Tony Finch's patch to make submission
1768 mode fix the return path as well as the Sender: header line, and to
1769 add a /name= option so that you can make the user's friendly name appear
1772 TF/03 Added the control = fakedefer ACL modifier.
1774 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1775 Mark Lowes for thorough testing.
1777 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1779 TK/03 Merged latest SRS patch from Miles Wilton.
1781 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1782 with the definition in sysexits.h (which is #included earlier).
1783 Fortunately, Exim does not actually use EX_OK. The code used to try to
1784 preserve the sysexits.h value, by assumimg that macro definitions were
1785 scanned for macro replacements. I have been disabused of this notion,
1786 so now the code just undefines EX_OK before #including unistd.h.
1788 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1789 in the smtp transport. When a block could not be written in a single
1790 write() function, the timeout was being re-applied to each part-write.
1791 This seems wrong - if the receiver was accepting one byte at a time it
1792 would take for ever. The timeout is now adjusted when this happens. It
1793 doesn't have to be particularly precise.
1795 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1796 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1798 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1799 without a domain if the check to <postmaster@domain> fails.
1801 SC/01 Eximstats: added -xls and the ability to specify output files
1802 (patch written by Frank Heydlauf).
1804 SC/02 Eximstats: use FileHandles for outputing results.
1806 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1808 SC/04 Eximstats: fixed display of large numbers with -nvr option
1810 SC/05 Eximstats: fixed merging of reports with empty tables.
1812 SC/06 Eximstats: added the -include_original_destination flag
1814 SC/07 Eximstats: removed tabs and trailing whitespace.
1816 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1818 TK/06 MBOX spool code: Add real "From " MBOX separator line
1819 so the .eml file is really in mbox format (even though
1820 most programs do not really care). Patch from Alex Miller.
1822 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1823 The latter is generated from $received_to and is only set if the
1824 message has one envelope recipient. SA can use these headers,
1825 obviously out-of-the-box. Patch from Alex Miller.
1827 PH/08 The ${def test on a variable was returning false if the variable's
1828 value was "0", contrary to what the specification has always said!
1829 The result should be true unless the variable is empty.
1831 PH/09 The syntax error of a character other than { following "${if
1832 def:variable_name" (after optional whitespace) was not being diagnosed.
1833 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1834 accidental colon was present, for example, could give incorrect results.
1836 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1837 result is used (not including appendfile, where other changes are about
1840 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1841 This involved changing a lot of size variables from int to off_t. It
1842 should work with maildirs and everything.
1844 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1845 spamd dying while we are connected to it.
1847 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1850 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1851 function for BDB 4.3.
1853 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1855 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1856 the sources for PCRE have been re-arranged and more files are now
1859 PH/15 The code I had for printing potentially long long variables in PH/11
1860 above was not the best (it lost precision). The length of off_t variables
1861 is now inspected at build time, and an appropriate printing format (%ld
1862 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1863 to be "long long int" or "long int". This is needed for the internal
1864 formatting function string_vformat().
1866 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1867 the configuration file to be ":syslog", then the script "guesses" where
1868 the logs files are, rather than using the compiled in default. In our
1869 case the guess is not the same as the compiled default, so the script
1870 suddenly stopped working when I started to use syslog. The patch checks
1871 to see if log_file_path is "". If so, it attempts to read it from exim
1872 with no configuration file to get the compiled in version, before it
1873 falls back to the previous guessing code."
1875 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1876 implementing BATV in an Exim configuration. See NewStuff for the gory
1879 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1880 Makefile that are specific to HP-UX.
1882 PH/18 If the "use_postmaster" option was set for a recipient callout together
1883 with the "random" option, the postmaster address was used as the MAIL
1884 FROM address for the random test, but not for the subsequent recipient
1885 test. It is now used for both.
1887 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1888 patch removes a few documentation additions to RFC 3028, because the
1889 latest draft now contains them. It adds the new en;ascii-case comparator
1890 and a new error check for 8bit text in MIME parts. Comparator and
1891 require names are now matched exactly. I enabled the subaddress
1892 extension, but it is not well tested yet (read: it works for me)."
1894 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1895 rework some of the code of TK/09 above to avoid the hardwired use of
1896 "%lld" and "long long". Replaced the call to snprintf() with a call to
1899 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1900 records point to non-existent hosts", "retry timeout exceeded", and
1901 "retry time not reached for any host after a long failure period".
1903 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1904 experimental DomainKeys support:
1906 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1907 (2) On an error such as an illegally used "control", the wrong name for
1908 the control was given.
1910 These problems did NOT occur unless DomainKeys support was compiled.
1912 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1914 PH/24 Added ${if match_ip condition.
1916 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1917 databases so that it will be absolutely obvious if a crash occurs in the
1918 DB library. This is a regular occurrence (often caused by mis-matched
1921 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1922 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1923 picked up on a user's system that detects such things. There doesn't seem
1924 to be a gcc warning option for this - only an attribute that has to be
1925 put on the function's prototype. It seems that in Fedora Core 4 they have
1926 set this on a number of new functions. No doubt there will be more in due
1929 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1930 specify a transport. However, if an address that was verified by such a
1931 router was the subject of a callout, Exim crashed because it tried to
1932 read the rcpt_include_affixes from the non-existent transport. Now it
1933 just assumes that the setting of that option is false. This bug was
1934 introduced by 4.51/PH/31.
1936 PH/28 Changed -d+all to exclude +memory, because that information is very
1937 rarely of interest, but it makes the output a lot bigger. People tend to
1938 do -d+all out of habit.
1940 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1941 code in os-type was giving problems when libc.so lives in lib64, like on
1944 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1945 aren't the modern standard, and it seems that some systems' include files
1946 don't always have them. Exim was already checking for some of the newer
1947 ones like T_AAAA, and defining it itself. I've added checks for all the
1948 record types that Exim uses.
1950 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1951 not automatically generating a new one, as it is supposed to. This
1952 prevented TLS from working. If the file did exist, but contained invalid
1953 data, a new version was generated, as expected. It was only the case of a
1954 non-existent file that was broken.
1956 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1957 with a change in libdomainkeys > 0.64.
1959 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1960 from DNS. If the selector record carries the flag, it now has
1961 precedence over the domain-wide flag.
1963 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1965 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1966 the use of a port name that isn't defined in /etc/services) occurred, the
1967 message was deferred as in a normal delivery, and thus remained on the
1968 spool, instead of being failed because of the mua_wrapper setting. This
1969 is now fixed, and I tidied up some of the mua_wrapper messages at the
1972 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1973 the messages in a hash of arrays rather than using individual hashes.
1974 This is a bit cleaner and results in dramatic memory savings, albeit
1975 at a slight CPU cost.
1977 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1978 as requested by Marc Sherman.
1980 SC/10 Eximstats: added histograms for user specified patterns as requested
1983 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1985 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1986 fopen() in the content-scanning modules that did not already have it.
1992 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1993 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1995 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1997 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1999 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2000 to test Sieve filters that use "vacation".
2002 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2003 that changes the way the GnuTLS parameters are stored in the cache file.
2004 The new format can be generated externally. For backward compatibility,
2005 if the data in the cache doesn't make sense, Exim assumes it has read an
2006 old-format file, and it generates new data and writes a new file. This
2007 means that you can't go back to an older release without removing the
2010 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2011 work if there are any delivery delays because "one_time" forces the
2012 parent to be marked "delivered", so its unseen clone is never tried
2013 again. For this reason, Exim now forbids the simultaneous setting of
2016 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2017 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2018 entry if you want to know the details. The fix, however, neglected to
2019 consider the case when local delivery batching is involved. The test for
2020 "previously delivered" was not happening when checking to see if an
2021 address could be batched with a previous (undelivered) one; under
2022 certain circumstances this could lead to multiple deliveries to the same
2025 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2026 in its include files, and this causes problems building Exim.
2028 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2029 header_syntax) but Exim was just ignoring anything given after a slash.
2030 In particular, this caused confusion with an attempt to use "verify =
2031 reverse_host_lookup/defer_ok". An error is now given when options are
2032 supplied for verify items that do not have them. (Maybe reverse_host_
2033 lookup should have a defer_ok option, but that's a different point.)
2035 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2036 defined by RFC 821) to 2048, because there were problems with some AUTH
2037 commands, and RFC 1869 says the size should be increased for extended
2038 SMTP commands that take arguments.
2040 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2043 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2044 "unknown" error; now it says that the functionality isn't in the binary.
2046 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2047 an address' error message when a string expansion fails (syntax or
2048 whatever). Otherwise the password may appear in the log. Following change
2049 PH/42 below, there is no longer a chance of it appearing in a bounce
2052 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2054 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2055 of its addresses was ever tried. (Bugzilla bug #2.)
2057 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2058 the result incorrectly in the debug output. (It correctly added a newline
2059 to what was transported.)
2061 TF/01 Added $received_time.
2063 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2064 commented out examples of how to interface to a virus scanner and to
2065 SpamAssassin. Also added commented examples of av_scanner and
2066 spamd_address settings.
2068 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2069 and controls are allowed in which ACLs. There were a couple of minor
2070 errors. Some of the entries in the conditions table (which is a table of
2071 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2072 negation of where the condition IS allowed.
2074 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2076 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2077 header file does not have a version number, so I've had to invent a new
2078 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2079 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2080 radiusclient), but it was contributed by a Radius user.
2082 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2083 files or directories, for OpenSSL.
2085 PH/19 When an Exim process that is running as root has to create an Exim log
2086 file, it does so in a subprocess that runs as exim:exim so as to get the
2087 ownership right at creation (otherwise, other Exim processes might see
2088 the file with the wrong ownership). There was no test for failure of this
2089 fork() call, which would lead to the process getting stuck as it waited
2090 for a non-existent subprocess. Forks do occasionally fail when resources
2091 run out. I reviewed all the other calls to fork(); they all seem to check
2094 PH/20 When checking for unexpected SMTP input at connect time (before writing
2095 the banner), Exim was not dealing correctly with a non-positive return
2096 from the read() function. If the client had disconnected by this time,
2097 the result was a log entry for a synchronization error with an empty
2098 string after "input=" when read() returned zero. If read() returned -1
2099 (an event I could not check), uninitialized data bytes were printed.
2100 There were reports of junk text (parts of files, etc) appearing after
2103 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2105 PH/22 Added support for macro redefinition, and (re)definition in between
2106 driver and ACL definitions.
2108 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2109 forgetting to use the resulting value; it was using the unexpanded value.
2111 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2112 hadn't been configured. The fix is from Juergen Kreileder, who
2113 understands it better than I do:
2115 "Here's what I see happening with three configured cyrus_sasl
2116 authenticators configured (plain, login, cram-md5):
2118 On startup auth_cyrus_sasl_init() gets called for each of these.
2119 This means three calls to sasl_listmech() without a specified mech_list.
2120 => SASL tests which mechs of all available mechs actually work
2121 => three warnings about OTP not working
2122 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2124 With the patch, sasl_listmech() also gets called three times. But now
2125 SASL's mech_list option is set to the server_mech specified in the the
2126 authenticator. Or in other words, the answer from sasl_listmech()
2127 gets limited to just the mech you're testing for (which is different
2129 => the return list contains just 'plain' or 'login', 'cram-md5' or
2130 nothing depending on the value of ob->server_mech.
2132 I've just tested the patch: Authentication still works fine,
2133 unavailable mechs specified in the exim configuration are still
2134 caught, and the auth.log warnings about OTP are gone."
2136 PH/25 When debugging is enabled, the contents of the command line are added
2137 to the debugging output, even when log_selector=+arguments is not
2140 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2141 answer is "GNU", and only if the return is "GNU/something" is the answer
2144 PH/27 $acl_verify_message is now set immediately after the failure of a
2145 verification in an ACL, and so is available in subsequent modifiers. In
2146 particular, the message can be preserved by coding like this:
2148 warn !verify = sender
2149 set acl_m0 = $acl_verify_message
2151 Previously, $acl_verify_message was set only while expanding "message"
2152 and "log_message" when a very denied access.
2154 PH/28 Modified OS/os.c-Linux with
2156 -#ifndef OS_LOAD_AVERAGE
2157 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2159 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2160 nomenclature these days.)
2162 PH/29 Installed patch from the Sieve maintainer that adds the options
2163 sieve_useraddress and sieve_subaddress to the redirect router.
2165 PH/30 In these circumstances:
2166 . Two addresses routed to the same list of hosts;
2167 . First host does not offer TLS;
2168 . First host accepts first address;
2169 . First host gives temporary error to second address;
2170 . Second host offers TLS and a TLS session is established;
2171 . Second host accepts second address.
2172 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2173 and peerdn, if requested) that were in fact used only for the second
2176 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2177 attention to any local part prefix or suffix that was matched by the
2178 router that accepted the address. It now behaves in the same way as it
2179 does for delivery: the affixes are removed from the local part unless
2180 rcpt_include_affixes is set on the transport.
2182 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2183 timeout during the DATA phase of an incoming message.
2185 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2186 applied a patch sanctioned by the Sieve maintainer.
2188 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2189 the uid or gid is negative. A case of a negative gid caused this to be
2190 noticed. The fix allows for either to be negative.
2192 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2193 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2196 PH/36 The change PH/12 above was broken. Fixed it.
2198 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2199 the grounds that routing the same address twice would always produce the
2200 same answer. This might have been true once, but it is certainly no
2201 longer true now. Routing a child address may depend on the previous
2202 routing that produced that child. Some complicated redirection strategies
2203 went wrong when messages had multiple recipients, and made Exim's
2204 behaviour dependent on the order in which the addresses were given.
2206 I have moved the duplicate checking until after the routing is complete.
2207 Exim scans the addresses that are assigned to local and remote
2208 transports, and removes any duplicates. This means that more work will be
2209 done, as duplicates will always all be routed, but duplicates are
2210 presumably rare, so I don't expect this is of any significance.
2212 For deliveries to pipes, files, and autoreplies, the duplicate checking
2213 still happens during the routing process, since they are not going to be
2216 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2217 It corrects a timeout issue with spamd. This is Ian's comment: "The
2218 background is that sometimes spamd either never reads data from a
2219 connection it has accepted, or it never writes response data. The exiscan
2220 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2221 blindly assumes that writes won't block so it may never time out."
2223 PH/39 Allow G after quota size as well as K and M.
2225 PH/40 The value set for $authenticated_id in an authenticator may not contain
2226 binary zeroes or newlines because the value is written to log lines and
2227 to spool files. There was no check on this. Now the value is run through
2228 the string_printing() function so that such characters are converted to
2229 printable escape sequences.
2231 PH/41 $message_linecount is a new variable that contains the total number of
2232 lines in the message. Compare $body_linecount, which is the count for the
2235 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2236 bounce and delay warning messages, except in certain special cases, which
2239 (a) An SMTP error message from a remote host;
2240 (b) A message specified in a :fail: redirection;
2241 (c) A message specified in a "fail" command in a system filter;
2242 (d) A message specified in a FAIL return from the queryprogram router;
2243 (e) A message specified by the cannot_route_message router option.
2245 In these cases only, Exim does include the error details in bounce and
2246 warning messages. There are also a few cases where bland messages such
2247 as "unrouteable address" or "local delivery error" are given.
2249 PH/43 $value is now also set for the "else" part of a ${run expansion.
2251 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2252 being worked on, but at least Exim now implements the latest version to
2255 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2256 process to complete was treated as a delivery failure, a timeout while
2257 writing the message to the pipe was logged, but erroneously treated as a
2258 successful delivery. Such timeouts include transport filter timeouts. For
2259 consistency with the overall process timeout, these timeouts are now
2260 treated as errors, giving rise to delivery failures by default. However,
2261 there is now a new Boolean option for the pipe transport called
2262 timeout_defer, which, if set TRUE, converts the failures into defers for
2263 both kinds of timeout. A transport filter timeout is now identified in
2266 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2267 systems where "make" and "gmake" are different, calling "gmake" at top
2268 level broke things. I've arranged for the value of $(MAKE) to be passed
2269 from the Makefile to this script so that it can call the same version of
2273 A note about Exim versions 4.44 and 4.50
2274 ----------------------------------------
2276 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2277 changes of various kinds. As a consequence, a big documentation update was
2278 needed. This delayed the release for rather longer than seemed good, especially
2279 in the light of a couple of (minor) security issues. Therefore, the changes
2280 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2281 release. So 4.44 and 4.50 are in effect two different branches that both start
2284 I have left the 4.50 change log unchanged; it contains all the changes since
2285 4.43. The change log for 4.44 is below; many of its items are identical to
2286 those for 4.50. This seems to be the most sensible way to preserve the
2287 historical information.
2293 1. Minor wording change to the doc/README.SIEVE file.
2295 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2296 computation of the current number of files was incorrect.
2298 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2299 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2302 4. Give more explanation in the error message when the command for a transport
2303 filter fails to execute.
2305 5. There are several places where Exim runs a non-Exim command in a
2306 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2307 was being done only for the command run by the queryprogram router. It is
2308 now done for all such subprocesses. The other cases are: ${run, transport
2309 filters, and the commands run by the lmtp and pipe transports.
2311 6. Added CONFIGURE_GROUP build-time option.
2313 7. Some older OS have a limit of 256 on the maximum number of file
2314 descriptors. Exim was using setrlimit() to set 1000 as a large value
2315 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2316 systems. I've change it so that if it can't get 1000, it tries for 256.
2318 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2319 was an oversight, and furthermore, ever since the addition of extra
2320 controls (e.g. 4.43/32), the checks on when to allow different forms of
2321 "control" were broken. There should now be diagnostics for all cases when a
2322 control that does not make sense is encountered.
2324 9. Added the /retain_sender option to "control=submission".
2326 10. $recipients is now available in the predata ACL (oversight).
2328 11. Tidy the search cache before the fork to do a delivery from a message
2329 received from the command line. Otherwise the child will trigger a lookup
2330 failure and thereby defer the delivery if it tries to use (for example) a
2331 cached ldap connection that the parent has called unbind on.
2333 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2334 of $address_data from the recipient verification was clobbered by the
2335 sender verification.
2337 13. The value of address_data from a sender verification is now available in
2338 $sender_address_data in subsequent conditions in the ACL statement.
2340 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2342 15. Added a new option "connect=<time>" to callout options, to set a different
2345 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2346 was its contents. (It was OK if the option was not defined at all.)
2348 17. A "Completed" log line is now written for messages that are removed from
2349 the spool by the -Mrm option.
2351 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2352 information about exactly what failed.
2354 19. Added -dd to debug only the daemon process.
2356 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2357 handles timeouts, both on the server side and network timeouts. Renamed the
2358 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2360 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2361 It is now set to "smtps".
2363 22. $host_address is now set to the target address during the checking of
2364 ignore_target_hosts.
2366 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2367 being passed; this would have caused $sender_host_name to have been used if
2368 matching the list had actually called for a host name (not very likely,
2369 since this list is usually IP addresses). A host name is now passed as
2372 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2373 code that specifies a non-restarting handler (typically sigaction() in
2374 modern systems) in an attempt to fix a rare and obscure crash bug.
2376 25. Narrowed the window for a race in the daemon that could cause it to ignore
2377 SIGCHLD signals. This is not a major problem, because they are used only to
2378 wake it up if nothing else does.
2380 26. A malformed maildirsize file could cause Exim to calculate negative values
2381 for the mailbox size or file count. Odd effects could occur as a result.
2382 The maildirsize information is now recalculated if the size or filecount
2385 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2386 support for a long time. Removed HAVE_SYS_VFS_H.
2388 28. Installed the latest version of exipick from John Jetmore.
2390 29. In an address list, if the pattern was not a regular expression, an empty
2391 subject address (from a bounce message) matched only if the pattern was an
2392 empty string. Non-empty patterns were not even tested. This was the wrong
2393 because it is perfectly reasonable to use an empty address as part of a
2394 database query. An empty address is now tested by patterns that are
2395 lookups. However, all the other forms of pattern expect the subject to
2396 contain a local part and a domain, and therefore, for them, an empty
2397 address still always fails if the pattern is not itself empty.
2399 30. Exim went into a mad DNS loop when attempting to do a callout where the
2400 host was specified on an smtp transport, and looking it up yielded more
2401 than one IP address.
2403 31. Re-factored the code for checking spool and log partition space into a
2404 function that finds that data and another that does the check. The former
2405 is then used to implement four new variables: $spool_space, $log_space,
2406 $spool_inodes, and $log_inodes.
2408 32. The RFC2047 encoding function was originally intended for short strings
2409 such as real names; it was not keeping to the 75-character limit for
2410 encoded words that the RFC imposes. It now respects the limit, and
2411 generates multiple encoded words if necessary. To be on the safe side, I
2412 have increased the buffer size for the ${rfc2047: expansion operator from
2415 33. It is now permitted to omit both strings after an "if" condition; if the
2416 condition is true, the result is "true". As before, when the second string
2417 is omitted, a false condition yields an empty string. This makes it less
2418 cumbersome to write custom ACL and router conditions.
2420 34. Failure to deliver a bounce message always caused it to be frozen, even if
2421 there was an errors_to setting on the router. The errors_to setting is now
2424 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2425 canonical form (fully expanded) before being placed in
2426 $sender_host_address.
2428 36. The table in the code that translates DNS record types into text (T_A to
2429 "A" for instance) was missing entries for NS and CNAME. It is just possible
2430 that this could have caused confusion if both these types were looked up
2431 for the same domain, because the text type is used as part of Exim's
2432 per-process caching. But the chance of anyone hitting this buglet seems
2435 37. The dnsdb lookup has been extended in a number of ways.
2437 (1) There is a new type, "zns", which walks up the domain tree until it
2438 finds some nameserver records. It should be used with care.
2440 (2) There is a new type, "mxh", which is like "mx" except that it returns
2441 just the host names, not the priorities.
2443 (3) It is now possible to give a list of domains (or IP addresses) to be
2444 looked up. The behaviour when one of the lookups defers can be
2445 controlled by a keyword.
2447 (4) It is now possible to specify the separator character for use when
2448 multiple records are returned.
2450 38. The dnslists ACL condition has been extended: it is now possible to supply
2451 a list of IP addresses and/or domains to be looked up in a particular DNS
2454 39. Added log_selector=+queue_time_overall.
2456 40. When running the queue in the test harness, wait just a tad after forking a
2457 delivery process, to get repeatability of debugging output.
2459 41. Include certificate and key file names in error message when GnuTLS fails
2460 to set them up, because the GnuTLS error message doesn't include the name
2461 of the failing file when there is a problem reading it.
2463 42. Allow both -bf and -bF in the same test run.
2465 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2467 44. The "Exiscan patch" is now merged into the mainline Exim source.
2469 45. Sometimes the final signoff response after QUIT could fail to get
2470 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2471 < 0 before doing a fflush(). This bug looks as though it goes back to the
2472 introduction of TLS in release 3.20, but "sometimes" must have been rare
2473 because the tests only now provoked it.
2475 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2476 (this can affect the format of dates).
2478 47. exim_tidydb, when checking for the continued existence of a message for
2479 which it has found a message-specific retry record, was not finding
2480 messages that were in split spool directories. Consequently, it was
2481 deleting retry records that should have stayed in existence.
2483 48. Steve fixed some bugs in eximstats.
2485 49. The SPA authentication driver was not abandoning authentication and moving
2486 on to the next authenticator when an expansion was forced to fail,
2487 contradicting the general specification for all authenticators. Instead it
2488 was generating a temporary error. It now behaves as specified.
2490 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2491 (the order specifies the preference for clients). The order is now AES256,
2492 AES128, 3DES, ARCFOUR128.
2494 51. Small patch to Sieve code - explicitly set From: when generating an
2497 52. Exim crashed if a remote delivery caused a very long error message to be
2498 recorded - for instance if somebody sent an entire SpamAssassin report back
2499 as a large number of 550 error lines. This bug was coincidentally fixed by
2500 increasing the size of one of Exim's internal buffers (big_buffer) that
2501 happened as part of the Exiscan merge. However, to be on the safe side, I
2502 have made the code more robust (and fixed the comments that describe what
2505 53. Now that there can be additional text after "Completed" in log lines (if
2506 the queue_time_overall log selector is set), a one-byte patch to exigrep
2507 was needed to allow it to recognize "Completed" as not the last thing in
2510 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2511 patch that reportedly fixes this has been added. I am not expert enough to
2512 create a test for it. This is what the patch creator wrote:
2514 "I found a little strange behaviour of ldap code when working with
2515 Windows 2003 AD Domain, where users was placed in more than one
2516 Organization Units. When I tried to give exim partial DN, the exit code
2517 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2518 But simultaneously result of request was absolutely normal ldap result,
2519 so I produce this patch..."
2521 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2522 REFERENCE, so I have modified the code to exclude the patch when that macro
2525 55. Some experimental protocols are using DNS PTR records for new purposes. The
2526 keys for these records are domain names, not reversed IP addresses. The
2527 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2528 leaves it alone. Component reversal etc. now happens only for IP addresses.
2531 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2533 57. Double the size of the debug message buffer (to 2048) so that more of very
2534 long debug lines gets shown.
2536 58. The exicyclog utility now does better if the number of log files to keep
2537 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2539 59. Two changes related to the smtp_active_hostname option:
2541 (1) $smtp_active_hostname is now available as a variable.
2542 (2) The default for smtp_banner uses $smtp_active_hostname instead
2543 of $primary_hostname.
2545 60. The host_aton() function is supposed to be passed a string that is known
2546 to be a valid IP address. However, in the case of IPv6 addresses, it was
2547 not checking this. This is a hostage to fortune. Exim now panics and dies
2548 if the condition is not met. A case was found where this could be provoked
2549 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2550 components; fortuitously, this particular loophole had already been fixed
2551 by change 4.50/55 above.
2553 If there are any other similar loopholes, the new check in host_aton()
2554 itself should stop them being exploited. The report I received stated that
2555 data on the command line could provoke the exploit when Exim was running as
2556 exim, but did not say which command line option was involved. All I could
2557 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2558 running as the user.
2561 61. There was a buffer overflow vulnerability in the SPA authentication code
2562 (which came originally from the Samba project). I have added a test to the
2563 spa_base64_to_bits() function which I hope fixes it.
2566 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2567 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2569 63. The daemon start-up calls getloadavg() while still root for those OS that
2570 need the first call to be done as root, but it missed one case: when
2571 deliver_queue_load_max is set with deliver_drop_privilege. This is
2572 necessary for the benefit of the queue runner, because there is no re-exec
2573 when deliver_drop_privilege is set.
2575 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2576 This has been fixed.
2578 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2579 was in use, was not putting the data itself into the right store pool;
2580 consequently, it could be overwritten for a subsequent message in the same
2581 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2584 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2586 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2587 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2588 the function were treating the return as a boolean value, which happened to
2589 work because 0=false and not-0=true, but is not correct code.
2591 68. The host_aton() function was not handling scoped IPv6 addresses (those
2592 with, for example, "%eth0" on the end) correctly.
2594 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2595 negated items (that is, ~something) in unsigned ints. Some compilers
2596 apparently mutter when there is no cast.
2598 70. If an address verification called from an ACL failed, and did not produce a
2599 user-specific message (i.e. there was only a "system" message), nothing was
2600 put in $acl_verify_message. In this situation, it now puts the system
2603 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2604 session; change 4.31/43 added the unwanted input to the log line - except
2605 that it did not do this in the start of session case. It now does.
2607 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2608 This is wrong and can cause the other end to generate a synchronization
2609 error if it is another Exim or anything else that does the synchronization
2610 check. A QUIT command is no longer sent after a timeout.
2612 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2613 during host lookups.
2615 74. The defer_ok option of callout verification was not working if it was used
2616 when verifying addresses in header lines, that is, for this case:
2618 verify = header_sender/callout=defer_ok
2620 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2621 those file descriptors could be used for SMTP connections. If anything
2622 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2623 could be sent to the SMTP client, causing chaos. The daemon now opens
2624 stdin, stdout, and stderr to /dev/null when it puts itself into the
2627 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2628 log by default. The user can override this with suitable Perl magic.
2630 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2631 the log message when discard triggers, was not working for the DATA ACL or
2632 for the non-SMTP ACL.
2634 78. Error message wording change in sieve.c.
2636 79. If smtp_accept_max_per_host was set, the number of connections could be
2637 restricted to fewer than expected, because the daemon was trying to set up
2638 a new connection before checking whether the processes handling previous
2639 connections had finished. The check for completed processes is now done
2640 earlier. On busy systems, this bug wouldn't be noticed because something
2641 else would have woken the daemon, and it would have reaped the completed
2644 80. If a message was submitted locally by a user whose login name contained one
2645 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2646 It caused a spool format error. I have fixed the spool reading code. A
2647 related problem was that the "from" clause in the Received: line became
2648 illegal because of the space(s). It is now covered by ${quote_local_part.
2650 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2653 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2655 83. Patch from Sieve maintainer to fix unterminated string problem in
2656 "vacation" handling.
2658 84. Some minor changes to the Linux configuration files to help with other
2659 OS variants using glibc.
2661 85. One more patch for Sieve to update vacation handling to latest spec.
2664 ----------------------------------------------------
2665 See the note above about the 4.44 and 4.50 releases.
2666 ----------------------------------------------------
2672 1. Change 4.43/35 introduced a bug that caused file counts to be
2673 incorrectly computed when quota_filecount was set in an appendfile
2676 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2677 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2680 3. Give more explanation in the error message when the command for a transport
2681 filter fails to execute.
2683 4. There are several places where Exim runs a non-Exim command in a
2684 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2685 was being done only for the command run by the queryprogram router. It is
2686 now done for all such subprocesses. The other cases are: ${run, transport
2687 filters, and the commands run by the lmtp and pipe transports.
2689 5. Some older OS have a limit of 256 on the maximum number of file
2690 descriptors. Exim was using setrlimit() to set 1000 as a large value
2691 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2692 systems. I've change it so that if it can't get 1000, it tries for 256.
2694 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2695 was an oversight, and furthermore, ever since the addition of extra
2696 controls (e.g. 4.43/32), the checks on when to allow different forms of
2697 "control" were broken. There should now be diagnostics for all cases when a
2698 control that does not make sense is encountered.
2700 7. $recipients is now available in the predata ACL (oversight).
2702 8. Tidy the search cache before the fork to do a delivery from a message
2703 received from the command line. Otherwise the child will trigger a lookup
2704 failure and thereby defer the delivery if it tries to use (for example) a
2705 cached ldap connection that the parent has called unbind on.
2707 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2708 of $address_data from the recipient verification was clobbered by the
2709 sender verification.
2711 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2712 was its contents. (It was OK if the option was not defined at all.)
2714 11. A "Completed" log line is now written for messages that are removed from
2715 the spool by the -Mrm option.
2717 12. $host_address is now set to the target address during the checking of
2718 ignore_target_hosts.
2720 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2721 being passed; this would have caused $sender_host_name to have been used if
2722 matching the list had actually called for a host name (not very likely,
2723 since this list is usually IP addresses). A host name is now passed as
2726 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2727 code that specifies a non-restarting handler (typically sigaction() in
2728 modern systems) in an attempt to fix a rare and obscure crash bug.
2730 15. Narrowed the window for a race in the daemon that could cause it to ignore
2731 SIGCHLD signals. This is not a major problem, because they are used only to
2732 wake it up if nothing else does.
2734 16. A malformed maildirsize file could cause Exim to calculate negative values
2735 for the mailbox size or file count. Odd effects could occur as a result.
2736 The maildirsize information is now recalculated if the size or filecount
2739 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2740 support for a long time. Removed HAVE_SYS_VFS_H.
2742 18. Updated exipick to current release from John Jetmore.
2744 19. Allow an empty sender to be matched against a lookup in an address list.
2745 Previously the only cases considered were a regular expression, or an
2748 20. Exim went into a mad DNS lookup loop when doing a callout where the
2749 host was specified on the transport, if the DNS lookup yielded more than
2752 21. The RFC2047 encoding function was originally intended for short strings
2753 such as real names; it was not keeping to the 75-character limit for
2754 encoded words that the RFC imposes. It now respects the limit, and
2755 generates multiple encoded words if necessary. To be on the safe side, I
2756 have increased the buffer size for the ${rfc2047: expansion operator from
2759 22. Failure to deliver a bounce message always caused it to be frozen, even if
2760 there was an errors_to setting on the router. The errors_to setting is now
2763 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2764 canonical form (fully expanded) before being placed in
2765 $sender_host_address.
2767 24. Updated eximstats to version 1.33
2769 25. Include certificate and key file names in error message when GnuTLS fails
2770 to set them up, because the GnuTLS error message doesn't include the name
2771 of the failing file when there is a problem reading it.
2773 26. Expand error message when OpenSSL has problems setting up cert/key files.
2776 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2777 (this can affect the format of dates).
2779 28. exim_tidydb, when checking for the continued existence of a message for
2780 which it has found a message-specific retry record, was not finding
2781 messages that were in split spool directories. Consequently, it was
2782 deleting retry records that should have stayed in existence.
2784 29. eximstats updated to version 1.35
2785 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2786 1.35 - bugfix such that pie charts by volume are generated correctly
2788 30. The SPA authentication driver was not abandoning authentication and moving
2789 on to the next authenticator when an expansion was forced to fail,
2790 contradicting the general specification for all authenticators. Instead it
2791 was generating a temporary error. It now behaves as specified.
2793 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2794 (the order specifies the preference for clients). The order is now AES256,
2795 AES128, 3DES, ARCFOUR128.
2797 31. Small patch to Sieve code - explicitly set From: when generating an
2800 32. Exim crashed if a remote delivery caused a very long error message to be
2801 recorded - for instance if somebody sent an entire SpamAssassin report back
2802 as a large number of 550 error lines. This bug was coincidentally fixed by
2803 increasing the size of one of Exim's internal buffers (big_buffer) that
2804 happened as part of the Exiscan merge. However, to be on the safe side, I
2805 have made the code more robust (and fixed the comments that describe what
2808 33. Some experimental protocols are using DNS PTR records for new purposes. The
2809 keys for these records are domain names, not reversed IP addresses. The
2810 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2811 leaves it alone. Component reversal etc. now happens only for IP addresses.
2814 34. The host_aton() function is supposed to be passed a string that is known
2815 to be a valid IP address. However, in the case of IPv6 addresses, it was
2816 not checking this. This is a hostage to fortune. Exim now panics and dies
2817 if the condition is not met. A case was found where this could be provoked
2818 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2819 components; fortuitously, this particular loophole had already been fixed
2820 by change 4.50/55 or 4.44/33 above.
2822 If there are any other similar loopholes, the new check in host_aton()
2823 itself should stop them being exploited. The report I received stated that
2824 data on the command line could provoke the exploit when Exim was running as
2825 exim, but did not say which command line option was involved. All I could
2826 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2827 running as the user.
2830 35. There was a buffer overflow vulnerability in the SPA authentication code
2831 (which came originally from the Samba project). I have added a test to the
2832 spa_base64_to_bits() function which I hope fixes it.
2835 36. The daemon start-up calls getloadavg() while still root for those OS that
2836 need the first call to be done as root, but it missed one case: when
2837 deliver_queue_load_max is set with deliver_drop_privilege. This is
2838 necessary for the benefit of the queue runner, because there is no re-exec
2839 when deliver_drop_privilege is set.
2841 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2842 was in use, was not putting the data itself into the right store pool;
2843 consequently, it could be overwritten for a subsequent message in the same
2844 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2847 38. Sometimes the final signoff response after QUIT could fail to get
2848 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2849 < 0 before doing a fflush(). This bug looks as though it goes back to the
2850 introduction of TLS in release 3.20, but "sometimes" must have been rare
2851 because the tests only now provoked it.
2857 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2858 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2859 Now it returns an integer. A number of calls were still expecting a T/F
2860 return. Fortuitously, in all cases, the tests worked in OK situations,
2861 which is the norm. However, things would have gone wrong on any write
2862 failures on the smtp file descriptor. This function is used when sending
2863 messages over SMTP and also when doing verify callouts.
2865 2. When Exim is called to do synchronous delivery of a locally submitted
2866 message (the -odf or -odi options), it no longer closes stderr before doing
2869 3. Implemented the mua_wrapper option.
2871 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2873 5. Implemented the functions header_remove(), header_testname(),
2874 header_add_at_position(), and receive_remove_recipient(), and exported them
2877 6. If an ACL "warn" statement specified the addition of headers, Exim already
2878 inserted X-ACL-Warn: at the start if there was no header name. However, it
2879 was not making this test for the second and subsequent header lines if
2880 there were newlines in the string. This meant that an invalid header could
2881 be inserted if Exim was badly configured.
2883 7. Allow an ACL "warn" statement to add header lines at the start or after all
2884 the Received: headers, as well as at the end.
2886 8. Added the rcpt_4xx retry error code.
2888 9. Added postmaster_mailfrom=xxx to callout verification option.
2890 10. Added mailfrom=xxxx to the callout verification option, for verify=
2893 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2894 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2896 12. Inserted some casts to stop certain compilers warning when using pointer
2897 differences as field lengths or precisions in printf-type calls (mostly
2898 affecting debugging statements).
2900 13. Added optional readline() support for -be (dynamically loaded).
2902 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2903 same clock tick as a message's arrival, so that its received time was the
2904 same as the "first fail" time on the retry record, and that message
2905 remained on the queue past the ultimate address timeout, every queue runner
2906 would try a delivery (because it was past the ultimate address timeout) but
2907 after another failure, the ultimate address timeout, which should have then
2908 bounced the address, did not kick in. This was a "< instead of <=" error;
2909 in most cases the first failure would have been in the next clock tick
2910 after the received time, and all would be well.
2912 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2913 being recognized when the domain list was tested by the match_domain
2914 condition in an expansion string.
2916 16. Added the ${str2b64: operator.
2918 17. Exim was always calling setrlimit() to set a large limit for the number of
2919 processes, without checking whether the existing limit was already
2920 adequate. (It did check for the limit on file descriptors.) Furthermore,
2921 errors from getrlimit() and setrlimit() were being ignored. Now they are
2922 logged to the main and panic logs, but Exim does carry on, to try to do its
2923 job under whatever limits there are.
2925 18. Imported PCRE 5.0.
2927 19. Trivial typo in log message " temporarily refused connection" (the leading
2930 20. If the log selector return_path_on_delivery was set and an address was
2931 redirected to /dev/null, the delivery process crashed because it assumed
2932 that a return path would always be set for a "successful" delivery. In this
2933 case, the whole delivery is bypassed as an optimization, and therefore no
2936 21. Internal re-arrangement: the function for sending a challenge and reading
2937 a response while authentication was assuming a zero-terminated challenge
2938 string. It's now changed to take a pointer and a length, to allow for
2939 binary data in such strings.
2941 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2943 23. Exim was not respecting finduser_retries when seeking the login of the
2944 uid under which it was called; it was always trying 10 times. (The default
2945 setting of finduser_retries is zero.) Also, it was sleeping after the final
2946 failure, which is pointless.
2948 24. Implemented tls_on_connect_ports.
2950 25. Implemented acl_smtp_predata.
2952 26. If the domain in control=submission is set empty, Exim assumes that the
2953 authenticated id is a complete email address when it generates From: or
2954 Sender: header lines.
2956 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2957 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2958 chown and chgrp in /bin and hostname in /usr/bin.
2960 28. Exim was keeping the "process log" file open after each use, just as it
2961 does for the main log. This opens the possibility of it remaining open for
2962 long periods when the USR1 signal hits a daemon. Occasional processlog
2963 errors were reported, that could have been caused by this. Anyway, it seems
2964 much more sensible not to leave this file open at all, so that is what now
2967 29. The long-running daemon process does not normally write to the log once it
2968 has entered its main loop, and it closes the log before doing so. This is
2969 so that log files can straightforwardly be renamed and moved. However,
2970 there are a couple of unusual error situations where the daemon does write
2971 log entries, and I had neglected to close the log afterwards.
2973 30. The text of an SMTP error response that was received during a remote
2974 delivery was being truncated at 512 bytes. This is too short for some of
2975 the long messages that one sometimes sees. I've increased the limit to
2978 31. It is now possible to make retry rules that apply only when a message has a
2979 specific sender, in particular, an empty sender.
2981 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2982 it possible to be selective about when SMTP synchronization is enforced.
2984 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2986 32. Implemented hosts_connection_nolog.
2988 33. Added an ACL for QUIT.
2990 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2993 35. Added mailbox_size and mailbox_filecount to appendfile.
2995 36. Added control = no_multiline_responses to ACLs.
2997 37. There was a bug in the logic of the code that waits for the clock to tick
2998 in the case where the clock went backwards by a substantial amount such
2999 that the microsecond fraction of "now" was more than the microsecond
3000 fraction of "then" (but the whole seconds number was less).
3002 38. Added support for the libradius Radius client library this is found on
3003 FreeBSD (previously only the radiusclient library was supported).
3009 1. When certain lookups returned multiple values in the form name=value, the
3010 quoting of the values was not always being done properly. Specifically:
3011 (a) If the value started with a double quote, but contained no whitespace,
3013 (b) If the value contained whitespace other than a space character (i.e.
3014 tabs or newlines or carriage returns) it was not quoted.
3015 This fix has been applied to the mysql and pgsql lookups by writing a
3016 separate quoting function and calling it from the lookup code. The fix
3017 should probably also be applied to nisplus, ibase and oracle lookups, but
3018 since I cannot test any of those, I have not disturbed their existing code.
3020 2. A hit in the callout cache for a specific address caused a log line with no
3021 reason for rejecting RCPT. Now it says "Previous (cached) callout
3022 verification failure".
3024 3. There was an off-by-one bug in the queryprogram router. An over-long
3025 return line was truncated at 256 instead of 255 characters, thereby
3026 overflowing its buffer with the terminating zero. As well as fixing this, I
3027 have increased the buffer size to 1024 (and made a note to document this).
3029 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3030 when Exim is waiting for an SMTP response from a remote server, Exim
3031 restarts its select() call on the socket, thereby resetting its timeout.
3032 This is not a problem when such interrupts are rare. Somebody set up a cron
3033 job to run exiwhat every 2 minutes, which is less than the normal select()
3034 timeout (5 or 10 minutes). This meant that the select() timeout never
3035 kicked in because it was always reset. I have fixed this by comparing the
3036 time when an interrupt arrives with the time at the start of the first call
3037 to select(). If more time than the timeout has elapsed, the interrupt is
3038 treated as a timeout.
3040 5. Some internal re-factoring in preparation for the addition of Sieve
3041 extensions (by MH). In particular, the "personal" test is moved to a
3042 separate function, and given an option for scanning Cc: and Bcc: (which is
3043 not set for Exim filters).
3045 6. When Exim created an email address using the login of the caller as the
3046 local part (e.g. when creating a From: or Sender: header line), it was not
3047 quoting the local part when it contained special characters such as @.
3049 7. Installed new OpenBSD configuration files.
3051 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3052 try to make them clearer.
3054 9. Callout options, other than the timeout value, were being ignored when
3055 verifying sender addresses in header lines. For example, when using
3057 verify = header_sender/callout=no_cache
3059 the cache was (incorrectly) being used.
3061 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3062 only the Cygwin environment.
3064 11. When return_path_on_delivery was set as a log selector, if different remote
3065 addresses in the same message used different return paths and parallel
3066 remote delivery occurred, the wrong values would sometimes be logged.
3067 (Whenever a remote delivery process finished, the return path value from
3068 the most recently started remote delivery process was logged.)
3070 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3071 lines when AUTH and/or TLS are in use. This is the "received protocol"
3072 field. Exim used to use "asmtp" for authenticated SMTP, without any
3073 indication (in the protocol name) for TLS use. Now it follows the RFC and
3074 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3075 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3076 names appear in log lines as well as in Received: header lines.
3078 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3079 extensions, and comparison tests, and to fix some bugs.
3081 14. Changes to the "personal" filter test:
3083 (1) The test was buggy in that it was just doing the equivalent of
3084 "contains" tests on header lines. For example, if a user's address was
3085 anne@some.where, the "personal" test would incorrectly be true for
3087 To: susanne@some.where
3089 This test is now done by extracting each address from the header in turn,
3090 and checking the entire address. Other tests that are part of "personal"
3091 are now done using regular expressions (for example, to check local parts
3092 of addresses in From: header lines).
3094 (2) The list of non-personal local parts in From: addresses has been
3095 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3096 taken from the Sieve specification recommendations.
3098 (3) If the message contains any header line starting with "List-" it is
3099 treated as non-personal.
3101 (4) The test for "circular" in the Subject: header line has been removed
3102 because it now seems ill-conceived.
3104 15. Minor typos in src/EDITME comments corrected.
3106 16. Installed latest exipick from John Jetmore.
3108 17. If headers_add on a router specified a text string that was too long for
3109 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3110 of string_sprintf() is now avoided.
3112 18. $message_body_size was not set (it was always zero) when running the DATA
3113 ACL and the local_scan() function.
3115 19. For the "mail" command in an Exim filter, no default was being set for
3116 the once_repeat time, causing a random time value to be used if "once" was
3117 specified. (If the value happened to be <= 0, no repeat happened.) The
3118 default is now 0s, meaning "never repeat". The "vacation" command was OK
3119 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3120 (I found it when inspecting the code).
3122 20. There is now an overall timeout for performing a callout verification. It
3123 defaults to 4 times the callout timeout, which applies to individual SMTP
3124 commands during the callout. The overall timeout applies when there is more
3125 than one host that can be tried. The timeout is checked before trying the
3126 next host. This prevents very long delays if there are a large number of
3127 hosts and all are timing out (e.g. when the network connections are timing
3128 out). The value of the overall timeout can be changed by specifying an
3129 additional sub-option for "callout", called "maxwait". For example:
3131 verify = sender/callout=5s,maxwait=20s
3133 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3134 to the end before writing, but this should make it even safer).
3136 22. Exim was forgetting that it had advertised PIPELINING for the second and
3137 subsequent messages on an SMTP connection. It was also not resetting its
3138 memory on STARTTLS and an internal HELO.
3140 23. When Exim logs an SMTP synchronization error within a session, it now
3141 records whether PIPELINING has been advertised or not.
3143 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3144 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3145 rather than long int.
3147 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3149 26. Added the never_mail option to autoreply.
3155 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3156 crash if the getsockname() call failed; this can happen if a connection is
3157 closed very soon after it is established. The problem was simply in the
3158 order in which certain operations were done, causing Exim to try to write
3159 to the SMTP stream before it had set up the file descriptor. The bug has
3160 been fixed by making things happen in the correct order.
3166 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3167 before the connection was closed, thus losing the rejection response.
3169 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3170 some early Solaris releases, but causes trouble in current releases where
3171 socklen_t is defined.
3173 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3176 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3179 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3180 that releasing the top of it at the end releases what was used for sub-
3181 expansions (unless the block got too big). However, discard this block if
3182 the first thing is a variable or header, so that we can use its block when
3183 it is dynamic (useful for very large $message_headers, for example).
3185 6. Lookups now cache *every* query, not just the most recent. A new, separate
3186 store pool is used for this. It can be recovered when all lookup caches are
3187 flushed. Lookups now release memory at the end of their result strings.
3188 This has involved some general refactoring of the lookup sources.
3190 7. Some code has been added to the store_xxx() functions to reduce the amount
3191 of flapping under certain conditions.
3193 8. log_incoming_interface used to affect only the <= reception log lines. Now
3194 it causes the local interface and port to be added to several more SMTP log
3195 lines, for example "SMTP connection from", and rejection lines.
3197 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3199 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3201 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3202 could be overwritten at the end of the current message (or the start of a
3203 new message if it was set in a HELO ACL). The value is now preserved for
3204 the duration of the SMTP connection.
3206 12. If a transport had a headers_rewrite setting, and a matching header line
3207 contained an unqualified address, that address was qualified, even if it
3208 did not match any rewriting rules. The underlying bug was that the values
3209 of the flags that permit the existence of unqualified sender and recipient
3210 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3211 non-local messages, and by -bnq for local messages) were not being
3212 preserved with the message after it was received.
3214 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3215 "next input=" as part of the text comprising the host identity instead of
3216 the correct text. The code was using the same buffer for two different
3217 strings. However, depending on which order the printing function evaluated
3218 its arguments, the bug did not always show up. Under Linux, for example, my
3219 test suite worked just fine.
3221 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3222 doesn't work with some older versions of Perl. It has been changed to "my",
3223 which in any case is probably the better facility to use.
3225 15. A really picky compiler found some instances of statements for creating
3226 error messages that either had too many or two few arguments for the format
3229 16. The size of the buffer for calls to the DNS resolver has been increased
3230 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3231 for addresses that have a lot of PTR records. This alleviates a problem; it
3232 does not fully solve it.
3234 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3235 buffer now truncates the list and logs the incident, which is the same
3236 action as happens when Exim is looking up a host name and its aliases.
3237 Previously in this situation something unpredictable would happen;
3238 sometimes it was "internal error: store_reset failed".
3240 18. If a server dropped the connection unexpectedly when an Exim client was
3241 using GnuTLS and trying to read a response, the client delivery process
3242 crashed while trying to generate an error log message.
3244 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3245 string, for example:
3247 warn message = H1: something\nH2: something
3249 the text was added as a single header line from Exim's point of view
3250 though it ended up OK in the delivered message. However, searching for the
3251 second and subsequent header lines using $h_h2: did not work. This has been
3252 fixed. Similarly, if a system filter added multiple headers in this way,
3253 the routers could not see them.
3255 20. Expanded the error message when iplsearch is called with an invalid key to
3256 suggest using net-iplsearch in a host list.
3258 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3259 ACLs are no longer actually imposed (and a message to that effect is
3262 22. If a "gecos" field in a passwd entry contained escaped characters, in
3263 particular, if it contained a \" sequence, Exim got it wrong when building
3264 a From: or a Sender: header from that name. A second bug also caused
3265 incorrect handling when an unquoted " was present following a character
3266 that needed quoting.
3268 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3269 was not being matched caselessly.
3271 24. Arranged for all hyphens in the exim.8 source to be escaped with
3274 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3275 properly. Recipient callout cache records were still being keyed to include
3276 the sender, even when use_sender was set false. This led to far more
3277 callouts that were necessary. The sender is no longer included in the key
3278 when use_sender is false.
3280 26. Added "control = submission" modifier to ACLs.
3282 27. Added the ${base62d: operator to decode base 62 numbers.
3284 28. dnsdb lookups can now access SRV records.
3286 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3287 the configuration file.
3289 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3290 (-v) mode. This makes the output for a verbose queue run more intelligible.
3292 31. Added a use_postmaster feature to recipient callouts.
3294 32. Added the $body_zerocount variable, containing the number of binary zero
3295 bytes in the message body.
3297 33. The time of last modification of the "new" subdirectory is now used as the
3298 "mailbox time last read" when there is a quota error for a maildir
3301 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3303 35. Added +ignore_unknown as a special item in host lists.
3305 36. Code for decoding IPv6 addresses in host lists is now included, even if
3306 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3307 address was recognized as an IP address, but was then not correctly decoded
3308 into binary, causing unexpected and incorrect effects when compared with
3315 1. Very minor rewording of debugging text in manualroute to say "list of
3316 hosts" instead of "hostlist".
3318 2. If verify=header_syntax was set, and a header line with an unqualified
3319 address (no domain) and a large number of spaces between the end of the
3320 name and the colon was received, the reception process suffered a buffer
3321 overflow, and (when I tested it) crashed. This was caused by some obsolete
3322 code that should have been removed. The fix is to remove it!
3324 3. When running in the test harness, delay a bit after writing a bounce
3325 message to get a bit more predictability in the log output.
3327 4. Added a call to search_tidyup() just before forking a reception process. In
3328 theory, someone could use a lookup in the expansion of smtp_accept_max_
3329 per_host which, without the tidyup, could leave open a database connection.
3331 5. Added the variables $recipient_data and $sender_data which get set from a
3332 lookup success in an ACL "recipients" or "senders" condition, or a router
3333 "senders" option, similar to $domain_data and $local_part_data.
3335 6. Moved the writing of debug_print from before to after the "senders" test
3338 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3339 problems for message scanning, either using a data ACL, or using
3340 local_scan() because the Received: header was not generated till after they
3341 were called (in order to set the time as the time of reception completion).
3342 I have revised the way this works. The header is now generated after the
3343 body is received, but before the ACL or local_scan() are called. After they
3344 are run, the timestamp in the header is updated.
3350 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3351 before starting a queue runner without re-exec. This happened only when
3352 deliver_drop_privilege was set or when the Exim user was set to root. The
3353 effect of the bug was that timeouts during subsequent deliveries caused
3354 crashes instead of being properly handled. The handler is now left at its
3355 default (and expected) setting.
3357 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3358 message, again when deliver_drop_privilege is set or Exim is run as root.
3359 The bug described in (1) was not present in this case, but the tidying up
3360 of the other signals was missing. I have made the two cases consistent.
3362 3. The ignore_target_hosts setting on a manualroute router was being ignored
3363 for hosts that were looked up using the /MX notation.
3365 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3368 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3369 operated on the sender address. After changing the $sender_address to <>
3370 for the sender address verify, Exim was re-instated it as the original
3371 (before rewriting) address, but remembering that it had rewritten it, so it
3372 wasn't rewriting it again. This bug also had the effect of breaking the
3373 sender address verification caching when the sender address was rewritten.
3375 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3376 This has been changed so that if the ip literal address matches
3377 ignore_target_hosts, the router declines.
3379 7. Added expansion conditions match_domain, match_address, and match_local_
3380 part (NOT match_host).
3382 8. The placeholder for the Received: header didn't have a length field set.
3384 9. Added code to Exim itself and to exim_lock to test for a specific race
3385 condition that could lead to file corruption when using MBX delivery. The
3386 issue is with the lockfile that is created in /tmp. If this file is removed
3387 after a process has opened it but before that process has acquired a lock,
3388 there is the potential for a second process to recreate the file and also
3389 acquire a lock. This could lead to two Exim processes writing to the file
3390 at the same time. The added code performs the same test as UW imapd; it
3391 checks after acquiring the lock that its file descriptor still refers to
3392 the same named file.
3394 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3395 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3398 11. Added the smtp_active_hostname option. If used, this will typically be made
3399 to depend on the incoming interface address. Because $interface_address is
3400 not set up until the daemon has forked a reception process, error responses
3401 that can happen earlier (such as "too many connections") no longer contain
3404 12. If an expansion in a condition on a "warn" statement fails because a lookup
3405 defers, the "warn" statement is abandoned, and the next ACL statement is
3406 processed. Previously this caused the whole ACL to be aborted.
3408 13. Added the iplsearch lookup type.
3410 14. Added ident_timeout as a log selector.
3412 15. Added tls_certificate_verified as a log selector.
3414 16. Added a global option tls_require_ciphers (compare the smtp transport
3415 option of the same name). This controls incoming TLS connections.
3417 17. I finally figured out how to make tls_require_ciphers do a similar thing
3418 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3419 before starting the TLS session.
3421 18. Tabs are now shown as \t in -bP output.
3423 19. If the log selector return_path_on_delivery was set, Exim crashed when
3424 bouncing a message because it had too many Received: header lines.
3426 20. If two routers both had headers_remove settings, and the first one included
3427 a superfluous trailing colon, the final name in the first list and the
3428 first name in the second list were incorrectly joined into one item (with a
3429 colon in the middle).
3435 1. Added -C and -D options to the exinext utility, mainly to make it easier
3436 to include in the automated testing, but these could be helpful when
3437 multiple configurations are in use.
3439 2. The exinext utility was not formatting the output nicely when there was
3440 an alternate port involved in the retry record key, nor when there was a
3441 message id as well (for retries that were specific to a specific message
3442 and a specific host). It was also confused by IPv6 addresses, because of
3443 the additional colons they contain. I have fixed the IPv4 problem, and
3444 patched it up to do a reasonable job for IPv6.
3446 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3447 delivery, the log line now contains "pipelined" if PIPELINING was used.
3449 4. An SMTP transport process used to panic and die if the bind() call to set
3450 an explicit outgoing interface failed. This has been changed; it is now
3451 treated in the same way as a connect() failure.
3453 5. A reference to $sender_host_name in the part of a conditional expansion
3454 that was being skipped was still causing a DNS lookup. This no longer
3457 6. The def: expansion condition was not recognizing references to header lines
3458 that used bh_ and bheader_.
3460 7. Added the _cache feature to named lists.
3462 8. The code for checking quota_filecount in the appendfile transport was
3463 allowing one more file than it should have been.
3465 9. For compatibility with Sendmail, the command line option
3473 and sets the incoming protocol and host name (for trusted callers). The
3474 host name and its colon can be omitted when only the protocol is to be set.
3475 Note the Exim already has two private options, -pd and -ps, that refer to
3476 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3477 "s", but I don't think that's a major issue.
3479 10. A number of refactoring changes to the code, none of which should affect
3482 (a) The number of logging options was getting close to filling up the
3483 32-bit word that was used as a bit map. I have split them into two classes:
3484 those that are passed in the argument to log_write(), and those that are
3485 only ever tested independently outside of that function. These are now in
3486 separate 32-bit words, so there is plenty of room for expansion again.
3487 There is no change in the user interface or the logging behaviour.
3489 (b) When building, for example, log lines, the code previously used a
3490 macro that called string_cat() twice, in order to add two strings. This is
3491 not really sufficiently general. Furthermore, there was one instance where
3492 it was actually wrong because one of the argument was used twice, and in
3493 one call a function was used. (As it happened, calling the function twice
3494 did not affect the overall behaviour.) The macro has been replaced by a
3495 function that can join an arbitrary number of extra strings onto a growing
3498 (c) The code for expansion conditions now uses a table and a binary chop
3499 instead of a serial search (which was left over from when there were very
3500 few conditions). Also, it now recognizes conditions like "pam" even when
3501 the relevant support is not compiled in: a suitably worded error message is
3502 given if an attempt is made to use such a condition.
3504 11. Added ${time_interval:xxxxx}.
3506 12. A bug was causing one of the ddress fields not to be passed back correctly
3507 from remote delivery subprocesses. The field in question was not being
3508 subsequently used, so this caused to problems in practice.
3510 13. Added new log selectors queue_time and deliver_time.
3512 14. Might have fixed a bug in maildirsizefile handling that threw up
3513 "unexpected character" debug warnings, and recalculated the data
3514 unnecessarily. In any case, I expanded the warning message to give more
3517 15. Added the message "Restricted characters in address" to the statements in
3518 the default ACL that block characters like @ and % in local parts.
3520 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3521 Three changes have been made:
3523 (a) There was a serious bug; a negative response to MAIL caused the whole
3524 recipient domain to be cached as invalid, thereby blocking all messages
3525 to all local parts at the same domain, from all senders. This bug has
3526 been fixed. The domain is no longer cached after a negative response to
3527 MAIL if the sender used is not empty.
3529 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3532 (c) A new callout option, "use_sender" has been added for people who want
3533 the modified behaviour.
3539 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3542 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3543 indeed breaks things for older releases.
3545 3. Added additional logging to the case where there is a problem reading data
3546 from a filter that is running in a subprocess using a pipe, in order to
3547 try to track down a specific problem.
3549 4. Testing facility fudge: when running in the test harness and attempting
3550 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3551 getting "No route to host". Convert this to a timeout.
3553 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3556 6. Some OS don't have socklen_t but use size_t instead. This affects the
3557 fifth argument of getsockopt() amongst other things. This is now
3558 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3559 can be set for individual OS. I have set it for SunOS5, OSF1, and
3560 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3561 some earlier ones do not.
3563 7. Change 4.30/15 was not doing the test caselessly.
3565 8. The standard form for an IPv6 address literal was being rejected by address
3566 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3567 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3568 this, as well as the form without the "IPv6" on the front (but only when
3569 address literals are enabled, of course).
3571 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3573 10. Exim crashed if a message with an empty sender address specified by -f
3574 encountered a router with an errors_to setting. This could be provoked only
3575 by a command such as
3579 where an empty string was supplied; "<>" did not hit this bug.
3581 11. Installed PCRE release 4.5.
3583 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3584 remained set. It is now erased.
3586 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3587 times from message ids (which are base 36 rather than the normal 62).
3589 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3590 were being counted as actual protocol errors, and logged if the log
3591 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3592 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3593 and DATA following a set of rejected RCPTs do not count as protocol errors.
3594 In other words, Exim assumes they were pipelined, though this may not
3595 actually be the case. Of course, in all cases the client gets an
3596 appropriate error code.
3598 15. If a lookup fails in an ACL condition, a message about the failure may
3599 be available; it is used if testing the ACL cannot continue, because most
3600 such messages specify what the cause of the deferral is. However, some
3601 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3602 that caused an old message to be retained and used if a later statement
3603 caused a defer, replacing the real cause of the deferral.
3605 16. If an IP address had so many PTR records that the DNS lookup buffer
3606 was not large enough to hold them, Exim could crash while trying to process
3607 the truncated data. It now detects and logs this case.
3609 17. Further to 4.21/58, another change has been made: if (and only if) the
3610 first line of a message (the first header line) ends with CRLF, a bare LF
3611 in a subsequent header line has a space inserted after it, so as not to
3612 terminate the header.
3614 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3615 unnecessarily, used atoi() instead of strtol(), and didn't check the
3616 termination when getting file sizes from file names by regex.
3618 19. Completely re-implemented the support for maildirsize files, in the light
3619 of a number of problems with the previous contributed implementation
3620 (4.30/29). In particular:
3622 . If the quota is zero, the maildirsize file is maintained, but no quota is
3625 . If the maildir directory does not exist, it is created before any attempt
3626 to write a maildirsize file.
3628 . The quota value in the file is just a cache; if the quota is changed in
3629 the transport, the new value overrides.
3631 . A regular expression is available for excluding directories from the
3634 20. The autoreply transport checks the characters in options that define the
3635 message's headers; it allows continued headers, but it was checking with
3636 isspace() after an embedded newline instead of explicitly looking for a
3639 21. If all the "regular" hosts to which an address was routed had passed their
3640 expiry times, and had not reached their retry times, the address was
3641 bounced, even if fallback hosts were defined. Now Exim should go on to try
3644 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3645 equivalent code in the SMTP transport. Some hosts send humungous responses
3646 to HELO/EHLO, more than 1024 it seems.
3648 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3649 gives compiler warnings in some environments. I've now done it "properly",
3652 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3653 (because of problems with the built-in one) was declared to return uschar *
3654 instead of char *, causing compiler failure.
3656 25. Fixed a file descriptor leak when processing alias/forward files.
3658 26. Fixed a minor format string issue in dbfn.c.
3660 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3662 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3663 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3666 29. When a sender address is verified, it is cached, to save repeating the test
3667 when there is more than one recipient in a message. However, when the
3668 verification involves a callout, it is possible for different callout
3669 options to be set for different recipients. It is too complicated to keep
3670 track of this in the cache, so now Exim always runs a verification when a
3671 callout is required, relying on the callout cache for the optimization.
3672 The overhead is duplication of the address routing, but this should not be
3675 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3676 to be verified with callout=postmaster, and the main callout worked but the
3677 postmaster check failed, the verification correctly failed. However, if a
3678 subsequent RCPT command asked for sender verification *without* the
3679 postmaster check, incorrect caching caused this verification also to fail,
3682 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3683 it was not caching the DNS options (qualify_single, search_parents) that
3684 were used when the lookup failed. A subsequent lookup with different
3685 options therefore always gave the same answer, though there were cases
3686 where it should not have. (Example: a "domains = !$mx_any" option on a
3687 dnslookup router: the "domains" option is always processed without any
3688 widening, but the router might have qualify_single set.) Now Exim uses the
3689 cached value only when the same options are set.
3691 32. Added John Jetmore's "exipick" utility to the distribution.
3693 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3694 than a timeout (e.g. a certificate is required, and is not provided), an
3695 Exim server now closes the connection immediately. Previously it waited for
3696 the client to close - but if the client is SSL, it seems that they each
3697 wait for each other, leading to a delay before one of them times out.
3699 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3700 maintained 0.8.x compatibility because I don't think many are using it, and
3701 it is clearly obsolete.
3703 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3706 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3707 client certificate was expired. A simple patch fixes this, though I don't
3708 understand the full logic of why the verify callback is called multiple
3711 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3712 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3713 which causes problems with some clients (such as the Certicom SSL Plus
3714 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3715 disables the coutermeasure allowing Eudora to connect."
3717 38. Exim was not checking that a write() to a log file succeeded. This could
3718 lead to Bad Things if a log got too big, in particular if it hit a file
3719 size limit. Exim now panics and dies if it cannot write to a log file, just
3720 as it does if it cannot open a log file.
3722 39. Modified OS/Makefile-Linux so that it now contains
3724 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3726 The two -D definitions ensure that Exim is compiled with large file
3727 support, which makes it possible to handle log files that are bigger than
3730 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3731 instance) a domain was checked against a named list that involved a lookup,
3732 causing $domain_data to be set, then another domain was checked against the
3733 same list, then the first domain was re-checked, the value of $domain_data
3734 after the final check could be wrong. In particular, if the second check
3735 failed, it could be set empty. This bug probably also applied to
3738 41. The strip_trailing_dot option was not being applied to the address given
3739 with the -f command-line option.
3741 42. The code for reading a message's header from the spool was incrementing
3742 $received_count, but never initializing it. This meant that the value was
3743 incorrect (doubled) while delivering a message in the same process in which
3744 it was received. In the most common configuration of Exim, this never
3745 happens - a fresh exec is done - but it can happen when
3746 deliver_drop_privilege is set.
3748 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3749 it now includes up to 150 characters of the unexpected data in the log
3752 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3753 and building data strings. The size of both of these buffers was 10 000
3754 bytes - far larger than anybody would *ever* want, thought I. Needless to
3755 say, somebody hit the limit. I have increased the maximum line length to
3756 20 000 and the maximum data length of concatenated lines to 100 000. I have
3757 also fixed two bugs, because there was no checking on these buffers. Tsk,
3758 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3759 buffer is too small.
3761 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3762 lsearch lookups. Now it does.
3764 46. When parsing a route_list item in a manualroute router, a fixed-length
3765 buffer was used for the list of hosts. I made this 1024 bytes long,
3766 thinking that nobody would ever have a list of hosts that long. Wrong.
3767 Somebody had a whole pile of complicated expansion conditions, and the
3768 string was silently truncated, leading to an expansion error. It turns out
3769 that it is easier to change to an unlimited length (owing to other changes
3770 that have happened since this code was originally written) than to build
3771 structure for giving a limitation error. The length of the item that
3772 expands into the list of hosts is now unlimited.
3774 47. The lsearch lookup could not handle data where the length of text line was
3775 more than 4095 characters. Such lines were truncated, leading to shortened
3776 data being returned. It should now handle lines of any length.
3778 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3779 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3782 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3783 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3784 no longer happens in comment lines. A list of replacements is now placed
3785 at the head of all of the source files, except those whose only change is
3786 to replace PERL_COMMAND in the very first #! line.
3788 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3789 messages on the queue, with a bottom-up merge sort, using code contributed
3790 by Michael Haardt. This should make operations like -bp somewhat faster on
3791 large queues. It won't affect queue runners, except when queue_run_in_order
3794 51. Installed eximstats 1.31 in the distribution.
3796 52. Added support for SRV lookups to the dnslookup router.
3798 53. If an ACL referred to $message_body or $message_body_end, the value was not
3799 reset for any messages that followed in the same SMTP session.
3801 54. The store-handling optimization for building very long strings was not
3802 differentiating between the different store pools. I don't think this
3803 actually made any difference in practice, but I've tidied it.
3805 55. While running the routers to verify a sender address, $sender_address
3806 was still set to the sender address. This is wrong, because when routing to
3807 send a bounce to the sender, it would be empty. Therefore, I have changed
3808 it so that, while verifying a sender address, $sender_address is set to <>.
3809 (There is no change to what happens when verifying a recipient address.)
3811 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3812 target A or AAAA records (if not already returned) without resetting the
3813 qualify_single or search_parents options of the DNS resolver. These are
3814 inappropriate in this case because the targets of MX and SRV records must
3815 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3816 target that, when qualified, matched something in the local domain. These
3817 two options are now turned off when doing these lookups.
3819 57. It seems that at least some releases of Reiserfs (which does not have the
3820 concept of a fixed number of inodes) returns zero and not -1 for the
3821 number of available inodes. This interacted badly with check_spool_inodes,
3822 which assumed that -1 was the "no such thing" setting. What I have done is
3823 to check that the total number of inodes is greater than zero before doing
3824 the test of how many are available.
3826 58. When a "warn" ACL statement has a log_message modifier, the message is
3827 remembered, and not repeated. This is to avoid a lot of repetition when a
3828 message has many recipients that cause the same warning to be written.
3829 Howewer, Exim was preserving the list of already written lines for an
3830 entire SMTP session, which doesn't seem right. The memory is now reset if a
3831 new message is started.
3833 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3834 debugging output unless log_rewrite was also set.
3836 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3837 of (int)(handle) when we know that handle contains (void *)(-1).
3839 61. The Exim daemon panic-logs an error return when it closes the incoming
3840 connection. However "connection reset by peer" seems to be common, and
3841 isn't really an error worthy of noting specially, so that particular error
3844 62. When Exim is trying to find all the local interfaces, it used to panic and
3845 die if the ioctl to get the interface flags failed. However, it seems that
3846 on at least one OS (Solaris 9) it is possible to have an interface that is
3847 included in the list of interfaces, but for which you get a failure error
3848 for this call. This happens when the interface is not "plumbed" into a
3849 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3850 failure of the "get flags" call assumes that the interface is down.
3852 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3853 makes life easier for people who are doing arithmetic on fields extracted
3854 from dates, where you often get leading zeros that should not be
3855 interpreted as octal.
3857 64. Added qualify_domain to the redirect router, to override the global
3860 65. If a pathologically long header line contained very many addresses (the
3861 report of this problem mentioned 10 000) and each of them was rewritten,
3862 Exim could use up a very large amount of memory. (It kept on making new
3863 copies of the header line as it rewrote, and never released the old ones.)
3864 At the expense of a bit more processing, the header rewriting function has
3865 been changed so that it no longer eats memory in this way.
3867 66. The generation of the Received: header has been moved from the time that a
3868 message starts to be received, to the time that it finishes. The timestamp
3869 in the Received: header should now be very close to that of the <= log
3870 line. There are two side-effects of this change:
3872 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3873 logged header lines no longer include the local Received: line, because
3874 it has not yet been created. The same applies to a copy of the message
3875 that is returned to a non-SMTP sender when a message is rejected.
3877 (b) When a filter file is tested using -bf, no additional Received: header
3878 is added to the test message. After some thought, I decided that this
3881 This change does not affect the value of $received_for. It is still set
3882 after address rewriting, but before local_scan() is called.
3884 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3886 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3887 gave an unhelpful panic error message, and a defer error. I have managed to
3888 change this behaviour so that it now rejects any supplied certificate,
3889 which seems right, as the list of acceptable certificates is empty.
3891 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3892 gave an unhelpful defer error. I have not managed to make this reject any
3893 supplied certificates, but the error message it gives is "no certificate
3894 supplied", which is not helpful.
3896 70. exigrep's output now also includes lines that are not associated with any
3897 message, but which match the given pattern. Implemented by a patch from
3898 Martin Sluka, which also tidied up the Perl a bit.
3900 71. Recipient callout verification, like sender verification, was using <> in
3901 the MAIL FROM command. This isn't really the right thing, since the actual
3902 sender may affect whether the remote host accepts the recipient or not. I
3903 have changed it to use the actual sender in the callout; this means that
3904 the cache record is now keyed on a recipient/sender pair, not just the
3905 recipient address. There doesn't seem to be a real danger of callout loops,
3906 since a callout by the remote host to check the sender would use <>.
3907 [SEE ABOVE: changed after hitting problems.]
3909 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3910 temporary errors. However, in the case of such a code being given after
3911 the end of a data transmission (i.e. after ".") Exim was failing to write
3912 a retry record for the message. (Yes, there was some broken host that was
3913 actually sending 8xx at this point.)
3915 73. An unknown lookup type in a host list could cause Exim to panic-die when
3916 the list was checked. (An example that provoked this was putting <; in the
3917 middle of a list instead of at the start.) If this happened during a DATA
3918 ACL check, a -D file could be left lying around. This kind of configuration
3919 error no longer causes Exim to die; instead it causes a defer errror. The
3920 incident is still logged to the main and panic logs.
3922 74. Buglet left over from Exim 3 conversion. The message "too many messages
3923 in one connection" was written to the rejectlog but not the mainlog, except
3924 when address rewriting (yes!) was being logged.
3926 75. Added write_rejectlog option.
3928 76. When a system filter was run not as root (that is, when system_filter_user
3929 was set), the values of the $n variables were not being returned to the
3930 main process; thus, they were not subsequently available in the $sn
3933 77. Added +return_path_on_delivery log selector.
3935 78. A connection timeout was being treated differently from recipients deferred
3936 when testing hosts_max_try with a message that was older than the host's
3937 retry timeout. (The host should not be counted, thus allowing all hosts to
3938 be tried at least once before bouncing.) This may have been the cause of an
3939 occasionally reported bug whereby a message would remain on the queue
3940 longer than the retry timeout, but would be bounced if a delivery was
3941 forced. I say "may" because I never totally pinned down the problem;
3942 setting up timeout/retry tests is difficult. See also the next item.
3944 79. The ultimate address timeout was not being applied to errors that involved
3945 a combination of host plus message (for example, a timeout on a MAIL
3946 command). When an address resolved to a number of possible hosts, and they
3947 were not all tried for each delivery (e.g. because of hosts_max_try), a
3948 message could remain on the queue longer than the retry timeout.
3950 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3953 81. Fixed an obscure SMTP outgoing bug which required at least the following
3954 conditions: (a) there was another message waiting for the same server;
3955 (b) the server returned 5xx to all RCPT commands in the first message so
3956 that the message was not completed; (c) the server dropped the connection
3957 or gave a negative response to the RSET that Exim sends to abort the
3958 transaction. The observed case was a dropped connection after DATA that had
3959 been sent in pipelining mode. That is, the server had advertised PIPELINING
3960 but was not implementing it correctly. The effect of the bug was incorrect
3961 behaviour, such as trying another host, and this could lead to a crash.
3967 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3968 and daemon.c were passed as pointers to ints; they should have been
3969 pointers to socklen_t variables (which are typically unsigned ints).
3971 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3974 3. Fixed a really odd bug that affected only the testing scheme; patching a
3975 certain fixed string in the binary changed the value of another string that
3976 happened to be identical to the end of the original first string.
3978 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3979 name", it returns that address as the IP address. On some operating
3980 systems (e.g. Solaris), it also passes back the IP address string as the
3981 "host name". However, on others (e.g. Linux), it passes back an empty
3982 string. Exim wasn't checking for this, and was changing the host name to an
3983 empty string, assuming it had been canonicized.
3985 5. Although rare, it is permitted to have more than one PTR record for a given
3986 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3987 all the names associated with an address, because they do in Solaris.
3988 However, it seems that they do not in Linux for data that comes from the
3989 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3990 I found this out when I moved to a new Linux workstation and tried to run
3991 the Exim test suite.
3993 To get round this problem I have changed the code so that it now does its
3994 own call to the DNS to look up PTR records when searching for a host name.
3995 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3996 addresses that are only in /etc/hosts are still found.
3998 This behaviour is, however, controlled by an option called host_lookup_
3999 order, which defaults to "bydns:byaddr". If people want to use the other
4000 order, or indeed, just use one or the other means of lookup, they can
4001 specify it in this variable.
4003 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4004 some operating systems, this comes back from gethostbyaddr() as an empty
4005 string, and this is what Exim used to test for. However, it seems that in
4006 other systems, "." is yielded. Exim now tests for this case too.
4008 7. The values of check_spool_space and check_log_space are now held internally
4009 as a number of kilobytes instead of an absolute number of bytes. If a
4010 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4011 kilobyte. This means that much larger values can be stored.
4013 8. Exim monitor: an attempt to get the action menu when not actually pointing
4014 at a message produces an empty menu entitled "No message selected". This
4015 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4016 no entries in it ("Shell widget menu has zero width and/or height"). So I
4017 have added a single, blank menu entry in this case.
4019 9. Added ${quote_local_part.
4021 10. MIME decoding is now applied to the contents of Subject: header lines when
4024 11. Now that a reference to $sender_host_address automatically causes a reverse
4025 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4026 host lookup before query-style lookups in lists that might use this
4027 variable. This has therefore been abolished, and the "net-" prefix is no
4028 longer necessary for query-style lookups.
4030 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4031 have been a typo for LFLAGS, so it has been changed.
4033 13. The install script calls Exim with "-C /dev/null" in order to find the
4034 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4035 to be output. Howeve, since Exim outputs its version number before the
4036 error, it didn't break the script. It just looked ugly. I fixed this by
4037 always allowing "-C /dev/null" if the caller is root.
4039 14. Ignore overlarge ACL variable number when reading spool file - insurance
4040 against a later release with more variables having written the file.
4042 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4043 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4044 this, as well as the form without the "IPv6" on the front.
4046 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4047 OS/Makefile-Darwin file.
4049 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4050 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4052 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4053 variable. However, some OS use ints for inodes. I've added cast to long int
4054 to get rid of the compiler warning.
4056 19. I had forgotten to lock out "/../" in configuration file names when
4057 ALT_CONFIG_PREFIX was set.
4059 20. Routers used for verification do not need to specify transports. However,
4060 if such a router generated a host list, and callout was configured, Exim
4061 crashed, because it could not find a port number from the (non-existent)
4062 transport. It now assumes port 25 in this circumstance.
4064 21. Added the -t option to exigrep.
4066 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4067 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4068 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4069 them is set, LOOKUP_LSEARCH is forced.
4071 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4073 24. Added sender and host information to the "rejected by local_scan()" log
4074 line; previously there was no indication of these.
4076 25. Added .include_if_exists.
4078 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4079 Linux. It turns out that not all file systems support this. Apparently some
4080 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4081 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4082 not supported on the file descriptor, is now ignored when Exim is trying to
4083 sync a directory. This applies only to Linux.
4085 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4087 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4088 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4091 29. Added support for maildirsize files from supplied patch (modified a bit).
4093 30. The use of :fail: followed by an empty string could lead Exim to respond to
4094 sender verification failures with (e.g.):
4096 550 Verification failed for <xxx>
4097 550 Sender verify failed
4099 where the first response line was missing the '-' that indicates it is not
4100 the final line of the response.
4102 31. The loop for finding the name of the user that called Exim had a hardwired
4103 limit of 10; it now uses the value of finduser_retries, which is used for
4104 all other user lookups.
4106 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4109 33. Exim was neglecting to zero errno before one call of strtol() when
4110 expanding a string and expecting an integer value. On some systems this
4111 resulted in spurious "integer overflow" errors. Also, it was casting the
4112 result into an int without checking.
4114 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4115 did not work. The code looks as if it has *never* worked, though it appears
4116 to have been documented since at least releast 1.62. I have made it work.
4118 35. The "timeout_DNS" error in retry rules, also documented since at least
4119 1.62, also never worked. As it isn't clear exactly what this means, and
4120 clearly it isn't a major issue, I have abolished the feature by treating it
4121 as "timeout", and writing a warning to the main and panic logs.
4123 36. The display of retry rules for -brt wasn't always showing the error code
4126 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4127 timeout_connect_A, timeout_connect_MX.
4129 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4130 to the empty sender.
4132 39. The daemon was not analysing the content of -oX till after it had closed
4133 stderr and disconnected from the controlling terminal. This meant that any
4134 syntax errors were only noted on the panic log, and the return code from
4135 the command was 0. By re-arranging the code a little, I've made the
4136 decoding happen first, so such errors now appear on stderr, and the return
4137 code is 1. However, the actual setting up of the sockets still happens in
4138 the disconnected process, so errors there are still only recorded on the
4141 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4142 connections (as happens on some IP stacks) was logged at start up time as
4143 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4144 it from "IPv6 and IPv4", which means that two separate sockets are being
4147 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4148 says whether AF_INET or AF_INET6 was passed as an argument.
4150 42. Exiwhat output was messed up when time zones were included in log
4153 43. Exiwhat now gives more information about the daemon's listening ports,
4154 and whether -tls-on-connect was used.
4156 44. The "port" option of the smtp transport is now expanded.
4158 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4159 silently ignored. Now an error message is written to the main and panic
4162 46. There's a new ACL modifier called "logwrite" which writes to a log file
4163 as soon as it is encountered.
4165 47. Added $local_user_uid and $local_user_gid at routing time.
4167 48. Exim crashed when trying to verify a sender address that was being
4170 49. Exim was recognizing only a space character after ".include". It now also
4171 recognizes a tab character.
4173 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4174 extracting the relevant information from the specification. The man page no
4175 longer contains scrambled data for the -d option, and I've added a section
4176 at the front about calling Exim under different names.
4178 51. Added "extra_headers" argument to the "mail" command in filter files.
4180 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4183 53. Installed eximstats 1.29.
4185 54. Added transport_filter_timeout as a generic transport option.
4187 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4188 Cc: header lines. This was required by RFC 822, but it not required by RFC
4191 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4192 incoming messages that did not have them. Now it does so only if the
4193 message originates locally, that is, if there is no associated remote host
4194 address. When Resent- header lines are present, this applies to the Resent-
4195 lines rather than the non-Resent- lines.
4197 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4198 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4200 58. Messages for configuration errors now include the name of the main
4201 configuration file - useful now that there may be more than one file in a
4202 list (.included file names were always shown).
4204 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4205 for those rare installations that do not start the daemon as root or run it
4206 setuid root. I've cut out the call to initgroups() if the daemon is not
4209 60. The Exim user and group can now be bound into the binary as text strings
4210 that are looked up at the start of Exim's processing.
4212 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4214 62. Added $mailstore_basename variable.
4216 63. Installed patch to sieve.c from Michael Haardt.
4218 64. When Exim failed to open the panic log after failing to open the main log,
4219 the original message it was trying to log was written to stderr and debug
4220 output, but if they were not available (the usual case in production), it
4221 was lost. Now it is written to syslog before the two lines that record the
4222 failures to open the logs.
4224 65. Users' Exim filters run in subprocesses under the user's uid. It is
4225 possible for a "deliver" command or an alias in a "personal" command to
4226 provoke an address rewrite. If logging of address rewriting is configured,
4227 this fails because the process is not running as root or exim. There may be
4228 a better way of dealing with this, but for the moment (because 4.30 needs
4229 to be released), I have disabled address rewrite logging when running a
4230 filter in a non-root, non-exim process.
4236 1. The buildconfig auxiliary program wasn't quoting the value set for
4237 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4238 not defined. This bug was masked in 4.22 by the effect that was fixed in
4241 2. Some messages that were rejected after a message id was allocated were
4242 shown as "incomplete" by exigrep. It no longer does this for messages that
4243 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4245 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4246 have allow_domain_literals set, the ID did not get logged in the <= line.
4247 Domain literals are now always recognized in Message-ID: header lines.
4249 4. The first argument for a ${extract expansion item is the key name or field
4250 number. Leading and trailing spaces in this item were not being ignored,
4251 causing some misleading effects.
4253 5. When deliver_drop_privilege was set, single queue runner processes started
4254 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4255 same command in the process it spins off) were not dropping privilege.
4257 6. When the daemon running as "exim" started a queue runner, it always
4258 re-executed Exim in the spun-off process. This is a waste of effort when
4259 deliver_drop_privilege is set. The new process now just calls the
4260 queue-runner function directly.
4266 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4269 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4270 ignored. Though the use of -oP was forcing the writing of a pid file, it
4271 was always written to the default place.
4273 3. If the message "no IP address found for host xxxx" is generated during
4274 incoming verification, it is now followed by identification of the incoming
4275 connection (so you can more easily find what provoked it).
4277 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4279 5. Added some features to "harden" Exim a bit more against certain attacks:
4281 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4282 be put in Local/Makefile. This is like the never_users runtime option,
4283 but it cannot be overridden. The default setting is "root".
4285 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4286 prefix string with which any file named in a -C command line option
4289 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4290 is retained for -C and -D only if the caller of Exim is root. Without
4291 it, the exim user may also use -C and -D and retain privilege.
4293 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4294 command line option is disabled.
4296 6. Macro names set by the -D option must start with an upper case letter, just
4297 like macro names defined in the configuration file.
4299 7. Added "dereference=" facility to LDAP.
4301 8. Two instances of the typo "uknown" in the source files are fixed.
4303 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4304 the Configure-Makefile script screwed up while processing it.
4306 10. Incorporated PCRE 4.4.
4308 11. The SMTP synchronization check was not operating right at the start of an
4309 SMTP session. For example, it could not catch a HELO sent before the client
4310 waited for the greeting. There is now a check for outstanding input at the
4311 point when the greeting is written. Because of the duplex, asynchronous
4312 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4313 way, but not yet received, when the check is performed.
4315 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4316 on TCP/IP sockets, because this apparently causes some broken clients to
4319 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4320 unchanged) from the Cygwin maintainer.
4322 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4323 was supported instead of testing for mbx. Effectively a typo.
4325 15. The spa authenticator server code was not checking that the input it
4326 received was valid base64.
4328 16. The debug output line for the "set" modifier in ACLs was not showing the
4329 name of the variable that was being set.
4331 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4333 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4334 lookup on its own. Something else was needed to trigger the lookup. For
4335 example, a match in host_lookup or the need for a host name in a host list.
4336 Now, if $sender_host_name is referenced and the host name has not yet been
4337 looked up, a lookup is performed. If the lookup fails, the variable remains
4338 empty, and $host_lookup_failed is set to "1".
4340 19. Added "eqi" as a case-independent comparison operator.
4342 20. The saslauthd authentication condition could segfault if neither service
4343 nor realm was specified.
4345 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4346 error message that was logged was misleading, and incoming SMTP
4347 connections were dropped. The message is now more accurate, and temporary
4348 errors are given to SMTP connections.
4350 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4351 (except RSET) with a 421 error, until QUIT is received. However, it was
4352 failing to send a response to QUIT.
4354 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4355 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4356 ACL is now run after the helo_try_verify_hosts code.
4358 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4359 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4360 case-independent, so other case variants are also recognized). Apparently
4361 some systems use these upper case variants.
4363 25. If more than two messages were waiting for the same host, and a transport
4364 filter was specified for the transport, Exim sent two messages over the
4365 same TCP/IP connection, and then failed with "socket operation on non-
4366 socket" when it tried to send the third.
4368 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4370 27. The extern definition of crypt16() in expand.c was not being excluded when
4371 the OS had its own crypt16() function.
4373 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4374 as a preferred synonym for return_size_limit, both as an option and as an
4377 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4379 30. Changed the default configuration ACL to relax the local part checking rule
4380 for addresses that are not in any local domains. For these addresses,
4381 slashes and pipe symbols are allowed within local parts, but the sequence
4382 /../ is explicitly forbidden.
4384 31. SPA server authentication was not clearing the challenge buffer before
4387 32. log_message in a "warn" ACL statement was writing to the reject log as
4388 well as to the main log, which contradicts the documentation and doesn't
4389 seem right (because no rejection is happening). So I have stopped it.
4391 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4392 However, I am unable to do any testing of this.
4394 34. Fixed an infelicity in the appendfile transport. When checking directories
4395 for a mailbox, to see if any needed to be created, it was accidentally
4396 using path names with one or more superfluous leading slashes; tracing
4397 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4399 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4400 contents are added to the log line that is written for every discarded
4401 recipient. (Previously a log_message setting was ignored.)
4403 36. The ${quote: operator now quotes the string if it is empty.
4405 37. The install script runs exim in order to find its version number. If for
4406 some reason other than non-existence or emptiness, which it checks, it
4407 could not run './exim', it was installing it with an empty version number,
4408 i.e. as "exim-". This error state is now caught, and the installation is
4411 38. An argument was missing from the function that creates an error message
4412 when Exim fails to connect to the socket for saslauthd authentication.
4413 This could cause Exim to crash, or give a corrupted message.
4415 39. Added isip, isip4, and isip6 to ${if conditions.
4417 40. The ACL variables $acl_xx are now saved with the message, and can be
4418 accessed later in routers, transports, and filters.
4420 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4421 strings in the file are not string-expanded.
4423 42. If a MAIL command specified a SIZE value that was too large to fit into an
4424 int variable, the check against message_size_limit failed. Such values are
4425 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4426 day this will have to be increased, but I don't think I want to be around
4427 when emails are that large.
4434 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4435 iconv() is not standard in FreeBSD.
4437 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4438 IPv6 enabled. The observed symptom was a segmentation fault on return from
4439 the function os_common_find_running_interfaces() in src/os.c.
4441 3. In the check_special_case() function in daemon.c I had used "errno" as an
4442 argument name, which causes warnings on some systems. This was basically a
4443 typo, since it was named "eno" in the comments!
4445 4. The code that waits for the clock to tick (at a resolution of some fraction
4446 of a second) so as to ensure message-id uniqueness was always waiting for
4447 at least one whole tick, when it could have waited for less. [This is
4448 almost certainly not relevant at current processor speeds, where it is
4449 unlikely to ever wait at all. But we try to future-proof.]
4451 5. The function that sleeps for a time interval that includes fractions of a
4452 second contained a race. It did not block SIGALRM between setting the
4453 timer, and suspending (a couple of lines later). If the interval was short
4454 and the sigsuspend() was delayed until after it had expired, the suspension
4455 never ended. On busy systems this could lead to processes getting stuck for
4458 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4459 process, before it forks any delivery processes. The open lookup caching
4460 mechanism meant that the open file or database connection was passed into
4461 the delivery process. The problem was that delivery processes always tidy
4462 up cached lookup data. This could cause a problem for the next delivery
4463 process started by the queue runner, because the external queue runner
4464 process does not know about the closure. So the next delivery process
4465 still has data in the lookup cache. In the case of a file lookup, there was
4466 no problem because closing a file descriptor in a subprocess doesn't affect
4467 the parent. However, if the lookup was caching a connection to a database,
4468 the connection was closed, and the second delivery process was likely to
4469 see errors such as "PGSQL: query failed: server closed the connection
4470 unexpectedly". The problem has been fixed by closing all cached lookups
4471 in a queue runner before running a delivery process.
4473 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4474 seem to have the "const" qualifier which it has on other OS. I've
4477 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4478 *of the same type* (client or server) with the same public name that an
4479 error should be diagnosed.
4481 9. When Exim looked up a host name for an IP address, but failed to find the
4482 original IP address when looking up the host name (a safety check), it
4483 output the message "<ip address> does not match any IP for NULL", which was
4484 confusing, to say the least. The bug was that the host name should have
4485 appeared instead of "NULL".
4487 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4488 user that is built into the binary, and the -C or -D options is used, root
4489 privilege is dropped before the configuration file is read. In addition,
4490 logging is switched to stderr instead of the normal log files. If the
4491 configuration then re-defines the Exim user, the unprivileged environment
4492 is probably not what is expected, so Exim logs a panic warning message (but
4495 However, if deliver_drop_privilege is set, the unprivileged state may well
4496 be exactly what is intended, so the warning has been cut out in that case,
4497 and Exim is allowed to try to write to its normal log files.
4503 1. smtp_return_error_details was not giving details for temporary sender
4504 or receiver verification errors.
4506 2. Diagnose a configuration error if two authenticators have the same public
4509 3. Exim used not to create the message log file for a message until the first
4510 delivery attempt. This could be confusing when incoming messages were held
4511 for policy or load reasons. The message log file is now created at the time
4512 the message is received, and an initial "Received" line is written to it.
4514 4. The automatically generated man page for command line options had a minor
4515 bug that caused no ill effects; however, a more serious problem was that
4516 the procedure for building the man page automatically didn't always
4517 operate. Consequently, release 4.20 contains an out-of-date version. This
4518 shouldn't happen again.
4520 5. When building Exim with embedded Perl support, the script that builds the
4521 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4522 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4524 6. The freeze_tell option was not being used for messages that were frozen on
4525 arrival, either by an ACL or by local_scan().
4527 7. Added the smtp_incomplete_transaction log selector.
4529 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4530 was accepting AUTH without a new EHLO.
4532 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4533 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4536 10. Logging of TCP/IP connections (when configured) now happens in the main
4537 daemon process instead of the child process, so that the TCP/IP connection
4538 count is more accurate (but it can never be perfect).
4540 11. The use of "drop" in a nested ACL was not being handled correctly in the
4541 outer ACL. Now, if condition failure induced by the nested "drop" causes
4542 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4543 or "require"), the connection is dropped.
4545 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4546 that yield "discard" can now be used with an "accept" or a "discard" verb,
4547 but an error is generated for any others (because I can't see a useful way
4548 to define what should happen).
4550 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4551 are now processed in the same way as lines in the Exim configuration file.
4552 In particular, continuation lines are supported.
4554 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4556 15. Added -ti meaning -t -i.
4558 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4559 domains, and warn by logging if others are found.
4561 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4562 returned from GIFCONF when Exim is trying to find the list of interfaces on
4563 a host. The code in os.c has been modified to copy each ifreq to an aligned
4564 structure in all cases.
4566 Also, in some cases, the returned ifreq's were being copied to a 'struct
4567 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4568 means the last couple of bytes of an IPv6 address could be chopped if the
4569 ifreq contained only a normal sockaddr (14 bytes storage).
4571 18. Named domain lists were not supported in the hosts_treat_as_local option.
4572 An entry such as +xxxx was not recognized, and was treated as a literal
4575 19. Ensure that header lines added by a DATA ACL are included in the reject log
4576 if the ACL subsequently rejects the message.
4578 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4579 MD5 (which is deprecated).
4581 21. When testing a filter file using -bf, Exim was writing a message when it
4582 took the sender from a "From " line in the message, but it was not doing so
4583 when it took $return_path from a Return-Path: header line. It now does.
4585 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4586 with a valid header line field name (a series of printing characters
4587 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4589 23. Changed "disc" in the source to "disk" to conform to the documentation and
4590 the book and for uniformity.
4592 24. Ignore Sendmail's -Ooption=value command line item.
4594 25. When execve() failed while trying to run a command in a pipe transport,
4595 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4596 could be confused with a return value of 69 from the command itself. This
4597 has been changed to 127, the value the shell returns if it is asked to run
4598 a non-existent command. The wording for the related log line suggests a
4599 non-existent command as the problem.
4601 26. If received_header_text expands to an empty string, do not add a Received:
4602 header line to the message. (Well, it adds a token one on the spool, but
4603 marks it "old" so that it doesn't get used or transmitted.)
4605 27. Installed eximstats 1.28 (addition of -nt option).
4607 28. There was no check for failure on the call to getsockname() in the daemon
4608 code. This can fail if there is a shortage of resources on the system, with
4609 ENOMEM, for example. A temporary error is now given on failure.
4611 29. Contrary to the C standard, it seems that in some environments, the
4612 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4613 program. Exim now does this explicitly; it affects the formatting of
4614 timestamps using strftime().
4616 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4617 complaints. I've now initialized all the variables, to avoid this.
4619 32. Header lines added by a system filter were not being "seen" during
4620 transport-time rewrites.
4622 33. The info_callback() function passed to OpenSSL is set up with type void
4623 (*)(SSL *, int, int), as described somewhere. However, when calling the
4624 function (actually a macro) that sets it up, the type void(*)() is
4625 expected. I've put in a cast to prevent warnings from picky compilers.
4627 34. If a DNS black list lookup found a CNAME record, but there were no A
4628 records associated with the domain it pointed at, Exim crashed.
4630 35. If a DNS black list lookup returned more than one A record, Exim ignored
4631 all but the first. It now scans all returned addresses if a particular IP
4632 value is being sought. In this situation, the contents of the
4633 $dnslist_value variable are a list of all the addresses, separated by a
4636 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4637 to accept a host name and all its aliases if the forward lookup for any of
4638 them yielded the IP address of the incoming connection. Now it accepts only
4639 those names whose forward lookup yields the correct IP address. Any other
4640 names are discarded. This closes a loophole whereby a rogue DNS
4641 administrator could create reverse DNS records to break through a
4642 wildcarded host restriction in an ACL.
4644 37. If a user filter or a system filter that ran in a subprocess used any of
4645 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4646 the wrong values were passed to the pipe command ($thisaddress had the
4647 value of $0, $0 had the value of $1, etc). This bug was introduced by
4648 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4650 38. Improved the line breaking for long SMTP error messages from ACLs.
4651 Previously, if there was no break point between 40 and 75 characters, Exim
4652 left the rest of the message alone. Two changes have been made: (a) I've
4653 reduced the minimum length to 35 characters; (b) if it can't find a break
4654 point between 35 and 75 characters, it looks ahead and uses the first one
4655 that it finds. This may give the occasional overlong line, but at least the
4656 remaining text gets split now.
4658 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4659 file descriptors might be low, and that setting 1000 would always raise it.
4660 It turns out that in some environments, the limit is already over 1000 and
4661 that lowering it causes trouble. So now Exim takes care not to decrease it.
4663 40. When delivering a message, the value of $return_path is set to $sender_
4664 address at the start of routing (routers may change the value). By an
4665 oversight, this default was not being set up when an address was tested by
4666 -bt or -bv, which affected the outcome if any router or filter referred to
4669 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4670 log only when "message" or "log_message" are set. However, if one of the
4671 conditions was an address verification, or a call to a nested ACL, the
4672 messages generated by the underlying test were being passed through. This
4673 no longer happens. The underlying message is available in $acl_verify_
4674 message for both "message" and "log_message" expansions, so it can be
4675 passed through if needed.
4677 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4678 new expansion $bh_ to give the encoded byte string without charset
4679 translation. Translation happens only if iconv() is available; HAVE_ICONV
4680 indicates this at build time. HEADERS_CHARSET gives the charset to
4681 translate to; headers_charset can change it in the configuration, and
4682 "headers charset" can change it in an individual filter file.
4684 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4685 that creates RFC 2047 encoded "words" labels them as that charset instead
4686 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4687 expansion operator; (ii) when Exim creates a From: line for a local
4688 message; (iii) when a header line is rewritten to include a "phrase" part.
4690 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4691 buggy, causing it to skip the first lines of messages whose message ID
4692 ended in 'D'. This would not have bitten before Exim release 4.14, because
4693 message IDs were unlikely to end in 'D' before then. The effect was to have
4694 incorrect size information for certain domains.
4696 45. #include "config.h" was missing at the start of the crypt16.c module. This
4697 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4700 46. If there was a timeout during a "random" callout check, Exim treated it as
4701 a failure of the random address, and carried on sending RSET and the real
4702 address. If the delay was just some slowness somewhere, the response to the
4703 original RCPT would be taken as a response to RSET and so on, causing
4704 mayhem of various kinds.
4706 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4707 when I implemented it. It didn't allow for the fact that some option values
4708 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4709 the right test for positive values.
4711 48. Domain names in DNS records are case-independent. Exim always looks them up
4712 in lower case. Some resolvers return domain names in exactly the case they
4713 appear in the zone file, that is, they may contain uppercase letters. Not
4714 all resolvers do this - some return always lower case. Exim was treating a
4715 change of case by a resolver as a change of domain, similar to a widening
4716 of a domain abbreviation. This triggered its re-routing code and so it was
4717 trying to route what was effectively the same domain again. This normally
4718 caused routing to fail (because the router wouldn't handle the domain
4719 twice). Now Exim checks for this case specially, and just changes the
4720 casing of the domain that it ultimately uses when it transmits the message
4723 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4726 50. If a filter generated a file delivery with a non-absolute name (possible if
4727 no home directory exists for the router), the forbid_file option was not
4730 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4731 the existing equality matching.
4733 52. Exim was using ints instead of ino_t variables in some places where it was
4734 dealing with inode numbers.
4736 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4737 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4738 called TMPDIR, and if it finds it is different, it changes its value.
4740 54. The smtp_printf() function is now made available to local_scan() so
4741 additional output lines can be written before returning. There is also an
4742 smtp_fflush() function to enable the detection of a dropped connection.
4743 The variables smtp_input and smtp_batched_input are exported to
4746 55. Changed the default runtime configuration: the message "Unknown user"
4747 has been removed from the ACL, and instead placed on the localuser router,
4748 using the cannot_route_message feature. This means that any verification
4749 failures that generate their own messages won't get overridden. Similarly,
4750 the "Unrouteable address" message that was in the ACL for unverifiable
4751 relay addresses has also been removed.
4753 56. Added hosts_avoid_esmtp to the smtp transport.
4755 57. The exicyclog script was not checking for the esoteric option
4756 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4757 will work only if exicyclog is run under the appropriate euid.
4759 58. Following a discussion on the list, the rules by which Exim recognises line
4760 endings on incoming messages have been changed. The -dropcr and drop_cr
4761 options are now no-ops, retained only for backwards compatibility. The
4762 following line terminators are recognized: LF CRLF CR. However, special
4763 processing applies to CR:
4765 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4766 nor a local message in the state where . is a terminator.
4768 (ii) If a bare CR is encountered in a header line, an extra space is added
4769 after the line terminator so as not to end the header. The reasoning
4770 behind this is that bare CRs in header lines are most likely either
4771 to be mistakes, or people trying to play silly games.
4773 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4774 was being incorrectly given as 18 bytes larger than it should have been.
4775 This is a VOB (very old bug).
4777 60. This may never have affected anything current, but just in case it has:
4778 When the local host is found other than at the start of a list of hosts,
4779 the local host, those with the same MX, and any that follow, are discarded.
4780 When the list in question was part of a longer list of hosts, the following
4781 hosts (not currently being processed) were also being discarded. This no
4782 longer happens. I'm not sure if this situation could ever has previously
4785 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4788 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4789 header. This is something that is recommended in a new Internet Draft, and
4790 is something that is documented as being done by Sendmail. There are two
4791 possible values. For messages generated by the autoreply transport, Exim
4794 Auto-Submitted: auto-replied
4796 whereas for all other generated messages (e.g. bounces) it adds
4798 Auto-Submitted: auto-generated
4800 63. The "personal" condition in filters now includes a test for the
4801 Auto-Submitted: header. If it contains the string "auto-" the message it
4802 not considered personal.
4804 64. Added rcpt_include_affixes as a generic transport option.
4806 65. Added queue_only_override (default true).
4808 66. Added the syslog_duplication option.
4810 67. If what should have been the first header line of a message consisted of
4811 a space followed by a colon, Exim was mis-interpreting it as a header line.
4812 It isn't of course - it is syntactically invalid and should therefore be
4813 treated as the start of the message body. The misbehaviour could have
4814 caused a number of strange effects, including loss of data in subsequent
4815 header lines, and spool format errors.
4817 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4818 client host had authenticated. This control can now be exercised by an ACL
4819 for more flexibility.
4821 69. By default, callouts do not happen when testing with -bh. There is now a
4822 variant, -bhc, which does actually run the callout code, including
4823 consulting and updating the callout cache.
4825 70. Added support for saslauthd authentication, courtesy of Alexander
4828 71. If statvfs() failed on the spool or log directories while checking their
4829 size for availability, Exim confusingly gave the error "space shortage".
4830 Furthermore, in debugging mode it crashed with a floating point exception.
4831 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4832 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4833 serious problem, Exim now writes to the main and panic logs when this
4834 happens, with details of the failure. It then refuses to accept the
4835 incoming message, giving the message "spool directory problem" or "log
4836 directory problem" with a 421 code for SMTP messages.
4838 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4839 0, 1, and 2 exist, because some OS complain for execs without them (see
4840 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4841 descriptors. However, the code omitted to check that the open succeeded,
4842 causing mysterious errors if for some reason the permissions on /dev/null
4843 got screwed. Now Exim writes a message to the main and panic logs, and
4844 bombs out if it can't open /dev/null.
4846 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4847 interact so that it is all more flexible. It is supposed to remain
4848 backwards compatible. Also added extra_local_interfaces.
4850 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4851 to bomb out with an assertion failure - to the client this appears as a
4852 connection drop. This problem occurs in the part of the code that was taken
4853 from the Samba project. Fortunately, the assertion is in a very simple
4854 function, so I have fixed this by reproducing the function inline in the
4855 one place where it is called, and arranging for authentication to fail
4856 instead of killing the process with assert().
4858 75. The SPA client code was not working when the server requested OEM rather
4859 than Unicode encoding.
4861 76. Added code to make require_files with a specific uid setting more usable in
4862 the case where statting the file as root fails - usually a non-root-mounted
4863 NFS file system. When this happens and the failure is EACCES, Exim now
4864 forks a subprocess and does the per-uid checking as the relevant uid.
4866 77. Added process_log_path.
4868 78. If log_file_path was not explicitly set, a setting of check_log_space or
4869 check_log_inodes was ignored.
4871 79. If a space check for the spool or log partitions fails, the incident is now
4872 logged. Of course, in the latter case the data may get lost...
4874 80. Added the %p formatting code to string_format() so that it can be used to
4875 print addresses in debug_print(). Adjusted all the address printing in the
4876 debugging in store.c to use %p rather than %d.
4878 81. There was a concern that a line of code in smtp_in.c could overflow a
4879 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4880 initially expressed, the concern was not well-founded, because trailing
4881 spaces are removed early. However, if the trailing spaces were followed by
4882 a NULL, they did not get removed, so the overflow was possible. Two fixes
4885 (a) I re-wrote the offending code in a cleaner fashion.
4886 (b) If an incoming SMTP command contains a NULL character, it is rejected
4889 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4890 runs initgroups(), so that if the Exim user is in any additional groups,
4891 they will be used during message reception.
4897 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob