1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.535 2007/10/18 12:08:47 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
6 TK/01 Add preliminary DKIM support. Currently requires a forked version of
7 ALT-N's libdkim that I have put here:
8 http://duncanthrax.net/exim-experimental/
10 Note to Michael Haardt: I had to rename some vars in sieve.c. They
11 were called 'true' and it seems that C99 defines that as a reserved
12 keyword to be used with 'bool' variable types. That means you could
13 not include C99-style headers which use bools without triggering
14 build errors in sieve.c.
16 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
17 as mailq or other aliases. Changed the --help handling significantly
18 to do whats expected. exim_usage() emits usage/help information.
20 SC/01 Added the -bylocaldomain option to eximstats.
22 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
24 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
29 PH/01 Another patch from the Sieve maintainer.
31 PH/02 When an IPv6 address is converted to a string for single-key lookup
32 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
33 dots are used instead of colons so that keys in lsearch files need not
34 contain colons. This was done some time before quoting was made available
35 in lsearch files. However, iplsearch files do require colons in IPv6 keys
36 (notated using the quote facility) so as to distinguish them from IPv4
37 keys. This meant that lookups for IP addresses in host lists did not work
38 for iplsearch lookups.
40 This has been fixed by arranging for IPv6 addresses to be expressed with
41 colons if the lookup type is iplsearch. This is not incompatible, because
42 previously such lookups could never work.
44 The situation is now rather anomolous, since one *can* have colons in
45 ordinary lsearch keys. However, making the change in all cases is
46 incompatible and would probably break a number of configurations.
48 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
51 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
52 conversion specification without a maximum field width, thereby enabling
53 a rogue spamd server to cause a buffer overflow. While nobody in their
54 right mind would setup Exim to query an untrusted spamd server, an
55 attacker that gains access to a server running spamd could potentially
56 exploit this vulnerability to run arbitrary code as the Exim user.
58 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
59 $primary_hostname instead of what libspf2 thinks the hosts name is.
61 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
62 a directory entry by the name of the lookup key. Previously, if a
63 symlink pointed to a non-existing file or a file in a directory that
64 Exim lacked permissions to read, a lookup for a key matching that
65 symlink would fail. Now it is enough that a matching directory entry
66 exists, symlink or not. (Bugzilla 503.)
68 PH/03 The body_linecount and body_zerocount variables are now exported in the
71 PH/04 Added the $dnslist_matched variable.
73 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
74 This means they are set thereafter only if the connection becomes
77 PH/06 Added the client_condition to authenticators so that some can be skipped
78 by clients under certain conditions.
80 PH/07 The error message for a badly-placed control=no_multiline_responses left
81 "_responses" off the end of the name.
83 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
85 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
86 (without spaces) instead of just copying the configuration text.
88 PH/10 Added the /noupdate option to the ratelimit ACL condition.
90 PH/11 Added $max_received_linelength.
92 PH/12 Added +ignore_defer and +include_defer to host lists.
94 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
95 way in which PCRE > 7.0 is built.
97 PH/14 Implemented queue_only_load_latch.
99 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
100 MAIL command. The effect was to mangle the value on 64-bit systems.
102 PH/16 Another patch from the Sieve maintainer.
104 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
106 PH/18 If a system quota error occurred while trying to create the file for
107 a maildir delivery, the message "Mailbox is full" was not appended to the
108 bounce if the delivery eventually timed out. Change 4.67/27 below applied
109 only to a quota excession during the actual writing of the file.
111 PH/19 It seems that peer DN values may contain newlines (and other non-printing
112 characters?) which causes problems in log lines. The DN values are now
113 passed through string_printing() before being added to log lines.
115 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
116 and InterBase are left for another time.)
118 PH/21 Added message_body_newlines option.
120 PH/22 Guard against possible overflow in moan_check_errorcopy().
122 PH/23 POSIX allows open() to be a macro; guard against that.
124 PH/24 If the recipient of an error message contained an @ in the local part
125 (suitably quoted, of course), incorrect values were put in $domain and
126 $local_part during the evaluation of errors_copy.
132 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
133 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
136 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
137 issue a MAIL command.
139 PH/02 In an ACL statement such as
141 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
143 if a client was not listed at all, or was listed with a value other than
144 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
145 the condition was not true (as it should be), so access was not denied.
146 The bug was that the ! inversion was incorrectly passed on to the second
147 item. This has been fixed.
149 PH/03 Added additional dnslists conditions == and =& which are different from
150 = and & when the dns lookup returns more than one IP address.
152 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
153 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
155 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
156 FSYNC, which compiles an option called disable_fsync that allows for
157 bypassing fsync(). The documentation is heavily laced with warnings.
159 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
161 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
162 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
163 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
164 including adding "make clean"; (3) Added -fPIC when compiling the test
165 dynamically loaded module, to get rid of a warning.
167 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
168 message fails, move_frozen_messages = true and ignore_bounce_errors_after
169 = 0s. The bug is otherwise harmless.
171 PH/07 There was a bug in the dovecot authenticator such that the value of
172 $auth1 could be overwritten, and so not correctly preserved, after a
173 successful authentication. This usually meant that the value preserved by
174 the server_setid option was incorrect.
176 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
178 PH/09 Installed PCRE release 7.0.
180 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
181 run for batched SMTP input. It is now run at the start of every message
182 in the batch. While fixing this I discovered that the process information
183 (output by running exiwhat) was not always getting set for -bs and -bS
184 input. This is fixed, and it now also says "batched" for BSMTP.
186 PH/11 Added control=no_pipelining.
188 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
189 patch, slightly modified), and move the expansion of helo_data till after
190 the connection is made in the smtp transport (so it can use these
193 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
195 PH/14 Added log_selector = +pid.
197 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
199 PH/16 Add ${if forany and ${if forall.
201 PH/17 Added dsn_from option to vary the From: line in DSNs.
203 PH/18 Flush SMTP output before performing a callout, unless control =
204 no_callout_flush is set.
206 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
207 was true (the default) a successful delivery failed to delete the retry
208 item, thus causing premature timeout of the address. The bug is now
211 PH/20 Added hosts_avoid_pipelining to the smtp transport.
213 PH/21 Long custom messages for fakedefer and fakereject are now split up
214 into multiline reponses in the same way that messages for "deny" and
215 other ACL rejections are.
217 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
218 with slight modification.
220 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
221 draft, changing the syntax and factoring some duplicate code".
223 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
224 for deliveries of the second and subsequent messages over the same SMTP
227 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
228 ${reduce, with only minor "tidies".
230 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
232 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
233 expansion side effects.
235 PH/27 When a message times out after an over-quota error from an Exim-imposed
236 quota, the bounce message says "mailbox is full". This message was not
237 being given when it was a system quota that was exceeded. It now should
240 MH/03 Made $recipients available in local_scan(). local_scan() already has
241 better access to the recipient list through recipients_list[], but
242 $recipients can be useful in postmaster-provided expansion strings.
244 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
245 in the case of a MAIL command with additional options following the
246 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
247 were accidentally chopped off.
249 PH/29 SMTP synchronization checks are implemented when a command is read -
250 there is a check that no more input is waiting when there shouldn't be
251 any. However, for some commands, a delay in an ACL can mean that it is
252 some time before the response is written. In this time, more input might
253 arrive, invalidly. So now there are extra checks after an ACL has run for
254 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
255 pipelining has not been advertised.
257 PH/30 MH's patch to allow iscntrl() characters to be list separators.
259 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
260 returned in the SMTP response when smtp_return_error_details was false.
263 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
264 instead of the C I/O that was originally supplied, because problems were
267 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
268 Exim which did not show up earlier: it was assuming that a call to
269 SSL_CTX_set_info_callback() might give an error value. In fact, there is
270 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
271 was a macro that became an assignment, so it seemed to work. This has
272 changed to a proper function call with a void return, hence the compile
273 error. Exim's code has been fixed.
275 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
278 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
280 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
282 PH/37 If a message is not accepted after it has had an id assigned (e.g.
283 because it turns out to be too big or there is a timeout) there is no
284 "Completed" line in the log. When some messages of this type were
285 selected by exigrep, they were listed as "not completed". Others were
286 picked up by some special patterns. I have improved the selection
287 criteria to be more general.
289 PH/38 The host_find_failed option in the manualroute router can now be set
290 to "ignore", to completely ignore a host whose IP address cannot be
291 found. If all hosts are ignored, the behaviour is controlled by the new
292 host_all_ignored option.
294 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
295 homing or because of multiple MX records with /mx) generated more than
296 one IP address, and the following item turned out to be the local host,
297 all the secondary addresses of the first item were incorrectly removed
298 from the list, along with the local host and any following hosts (which
299 is what is supposed to happen).
301 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
302 whoever called Exim into the -H file. In the case of the daemon it was
303 behaving confusingly. When first started, it used values for whoever
304 started the daemon, but after a SIGHUP it used the Exim user (because it
305 calls itself on a restart). I have changed the code so that it now always
308 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
309 message are rejected with the same error (e.g. no authentication or bad
310 sender address), and a DATA command is nevertheless sent (as can happen
311 with PIPELINING or a stupid MUA), the error message that was given to the
312 RCPT commands is included in the rejection of the DATA command. This is
313 intended to be helpful for MUAs that show only the final error to their
316 PH/42 Another patch from the Sieve maintainer.
318 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
319 Eximstats - Fixed some broken HTML links and added missing column headers
321 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
322 columns for Rejects, Temp Rejects, Ham, and Spam rows.
324 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
326 PH/43 Yet another patch from the Sieve maintainer.
328 PH/44 I found a way to check for a TCP/IP connection going away before sending
329 the response to the final '.' that terminates a message, but only in the
330 case where the client has not sent further data following the '.'
331 (unfortunately, this is allowed). However, in many cases there won't be
332 any further data because there won't be any more messages to send. A call
333 to select() can be used: if it shows that the input is "ready", there is
334 either input waiting, or the socket has been closed. An attempt to read
335 the next input character can distinguish the two cases. Previously, Exim
336 would have sent an OK response which the client would never have see.
337 This could lead to message repetition. This fix should cure that, at
338 least in a lot of common cases.
340 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
341 advertised in response to EHLO.
347 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
348 fixed by 4.65/MH/01 (is this a record?) are fixed:
350 (i) An empty string was always treated as zero by the numeric comparison
351 operators. This behaviour has been restored.
353 (ii) It is documented that the numeric comparison operators always treat
354 their arguments as decimal numbers. This was broken in that numbers
355 starting with 0 were being interpreted as octal.
357 While fixing these problems I realized that there was another issue that
358 hadn't been noticed. Values of message_size_limit (both the global option
359 and the transport option) were treated as octal if they started with 0.
360 The documentation was vague. These values are now always treated as
361 decimal, and I will make that clear in the documentation.
367 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
368 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
371 MH/01 Don't check that the operands of numeric comparison operators are
372 integers when their expansion is in "skipping" mode (fixes bug
373 introduced by 4.64-PH/07).
375 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
376 child addresses, Exim now panics and dies. Previously, because the count
377 is held in a short int, deliveries were likely to be lost. As such a
378 large number of recipients for a single message is ridiculous
379 (performance will be very, very poor), I have chosen to impose a limit
380 rather than extend the field.
386 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
387 leftover -K file (the existence of which was triggered by #402).
388 While we were at it, introduced process PID as part of the -K
389 filename. This should rule out race conditions when creating
392 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
393 processing considerably. Previous code took too long for large mails,
394 triggering a timeout which in turn triggers #401.
396 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
397 in the DK code in transports.c. sendfile() is not really portable,
398 hence the _LINUX specificness.
400 TF/01 In the add_headers option to the mail command in an Exim filter,
401 there was a bug that Exim would claim a syntax error in any
402 header after the first one which had an odd number of characters
405 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
406 callout verification, Exim cached a "reject" for the entire domain. This
407 is correct for most verifications, but it is not correct for a recipient
408 verification with use_sender or use_postmaster set, because in that case
409 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
410 case of MAIL FROM:<> rejection from other early rejections (e.g.
411 rejection of HELO). When verifying a recipient using a non-null MAIL
412 address, the cache is ignored if it shows MAIL FROM:<> rejection.
413 Whatever the result of the callout, the value of the domain cache is
414 left unchanged (for any other kind of callout, getting as far as trying
415 RCPT means that the domain itself is ok).
417 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
420 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
421 manifest itself as EPIPE rather than ECONNECT. When tidying away a
422 session, the daemon ignores ECONNECT errors and logs others; it now
423 ignores EPIPE as well.
425 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
426 (quoted-printable decoding).
428 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
429 later the small subsequent patch to fix an introduced bug.
431 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
433 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
435 PH/08 An error is now given if message_size_limit is specified negative.
437 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
438 to be given (somewhat) arbitrary names.
440 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
443 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
444 miscellaneous code fixes
446 PH/10 Added the log_reject_target ACL modifier to specify where to log
449 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
450 hostname. This is wrong, because it relates to the incoming message (and
451 probably the interface on which it is arriving) and not to the outgoing
452 callout (which could be using a different interface). This has been
453 changed to use the value of the helo_data option from the smtp transport
454 instead - this is what is used when a message is actually being sent. If
455 there is no remote transport (possible with a router that sets up host
456 addresses), $smtp_active_hostname is used.
458 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
459 tweaks were necessary in order to get it to work (see also 21 below):
460 (a) The code assumed that strncpy() returns a negative number on buffer
461 overflow, which isn't the case. Replaced with Exim's string_format()
463 (b) There were several signed/unsigned issues. I just did the minimum
464 hacking in of casts. There is scope for a larger refactoring.
465 (c) The code used strcasecmp() which is not a standard C function.
466 Replaced with Exim's strcmpic() function.
467 (d) The code set only $1; it now sets $auth1 as well.
468 (e) A simple test gave the error "authentication client didn't specify
469 service in request". It would seem that Dovecot has changed its
470 interface. Fortunately there's a specification; I followed it and
471 changed what the client sends and it appears to be working now.
473 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
476 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
477 address A is aliased to B and C, where B exists and C does not. Without
478 -v the output is "A verified" because verification stops after a
479 successful redirection if more than one address is generated. However,
480 with -v the child addresses are also verified. Exim was outputting "A
481 failed to verify" and then showing the successful verification for C,
482 with its parentage. It now outputs "B failed to verify", showing B's
483 parentage before showing the successful verification of C.
485 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
486 look up a TXT record in a specific list after matching in a combined
489 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
490 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
491 they consult the DNS. I had assumed they would set it the way they
492 wanted; and indeed my experiments on Linux seem to show that in some
493 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
494 To be on the safe side, however, I have now made the interface to
495 host_find_byname() similar to host_find_bydns(), with an argument
496 containing the DNS resolver options. The host_find_byname() function now
497 sets these options at its start, just as host_find_bydns() does. The smtp
498 transport options dns_qualify_single and dns_search_parents are passed to
499 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
500 of host_find_byname() use the default settings of RES_DEFNAMES
501 (qualify_single) but not RES_DNSRCH (search_parents).
503 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
504 spool_read_header() do less string testing, by means of a preliminary
505 switch on the second character of optional "-foo" lines. (This is
506 overdue, caused by the large number of possibilities that now exist.
507 Originally there were few.) While I was there, I also converted the
508 str(n)cmp tests so they don't re-test the leading "-" and the first
509 character, in the hope this might squeeze out yet more improvement.
511 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
512 flag allowing group syntax was set by the header_syntax check but not
513 turned off, possible causing trouble later; (2) The flag was not being
514 set at all for the header_verify test, causing "group"-style headers to
515 be rejected. I have now set it in this case, and also caused header_
516 verify to ignore an empty address taken from a group. While doing this, I
517 came across some other cases where the code for allowing group syntax
518 while scanning a header line wasn't quite right (mostly, not resetting
519 the flag correctly in the right place). These bugs could have caused
520 trouble for malformed header lines. I hope it is now all correct.
522 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
523 with the "reply" argument non-NULL. The code, however (which originally
524 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
525 but it didn't always do it. This confused somebody who was copying the
526 code for some other use. I have removed all the tests.
528 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
529 feature that was used to support insecure browsers during the U.S. crypto
530 embargo. It requires special client support, and Exim is probably the
531 only MTA that supported it -- and would never use it because real RSA is
532 always available. This code has been removed, because it had the bad
533 effect of slowing Exim down by computing (never used) parameters for the
534 RSA_EXPORT functionality.
536 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
537 authenticator to fail if there's a tab character in the incoming data
538 (there should never be unless someone is messing about, as it's supposed
539 to be base64-encoded). Also added, on Timo's advice, the "secured" option
540 if the connection is using TLS or if the remote IP is the same as the
541 local IP, and the "valid-client-cert option" if a client certificate has
544 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
545 authenticators. This can be used for authorization after authentication
546 succeeds. (In the case of plaintext, it servers for both authentication
549 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
550 if any retry times were supplied.
552 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
553 connection, where there is no client IP address to check. In this
554 situation, the verify now always succeeds.
556 PH/25 Applied John Jetmore's -Mset patch.
558 PH/26 Added -bem to be like -Mset, but loading a message from a file.
560 PH/27 In a string expansion for a processed (not raw) header when multiple
561 headers of the same name were present, leading whitespace was being
562 removed from all of them, but trailing whitespace was being removed only
563 from the last one. Now trailing whitespace is removed from each header
564 before concatenation. Completely empty headers in a concatenation (as
567 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
568 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
570 PH/29 [Removed. This was a change that I later backed out, and forgot to
571 correct the ChangeLog entry (that I had efficiently created) before
572 committing the later change.]
574 PH/30 Exim was sometimes attempting to deliver messages that had suffered
575 address errors (4xx response to RCPT) over the same connection as other
576 messages routed to the same hosts. Such deliveries are always "forced",
577 so retry times are not inspected. This resulted in far too many retries
578 for the affected addresses. The effect occurred only when there were more
579 hosts than the hosts_max_try setting in the smtp transport when it had
580 the 4xx errors. Those hosts that it had tried were not added to the list
581 of hosts for which the message was waiting, so if all were tried, there
582 was no problem. Two fixes have been applied:
584 (i) If there are any address or message errors in an SMTP delivery, none
585 of the hosts (tried or untried) are now added to the list of hosts
586 for which the message is waiting, so the message should not be a
587 candidate for sending over the same connection that was used for a
588 successful delivery of some other message. This seems entirely
589 reasonable: after all the message is NOT "waiting for some host".
590 This is so "obvious" that I'm not sure why it wasn't done
591 previously. Hope I haven't missed anything, but it can't do any
592 harm, as the worst effect is to miss an optimization.
594 (ii) If, despite (i), such a delivery is accidentally attempted, the
595 routing retry time is respected, so at least it doesn't keep
596 hammering the server.
598 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
599 in ${readsocket because some servers need this prod.
601 PH/32 Added some extra debug output when updating a wait-xxx database.
603 PH/33 The hint "could be header name not terminated by colon", which has been
604 given for certain expansion errors for a long time, was not being given
605 for the ${if def:h_colon_omitted{... case.
607 PH/34 The spec says: "With one important exception, whenever a domain list is
608 being scanned, $domain contains the subject domain." There was at least
609 one case where this was not true.
611 PH/35 The error "getsockname() failed: connection reset by peer" was being
612 written to the panic log as well as the main log, but it isn't really
613 panic-worthy as it just means the connection died rather early on. I have
614 removed the panic log writing for the ECONNRESET error when getsockname()
617 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
618 runs only) independently of the message's sender address. This meant
619 that, if the 4xx error was in fact related to the sender, a different
620 message to the same recipient with a different sender could confuse
621 things. In particualar, this can happen when sending to a greylisting
622 server, but other circumstances could also provoke similar problems.
623 I have changed the default so that the retry time for these errors is now
624 based a combination of the sender and recipient addresses. This change
625 can be overridden by setting address_retry_include_sender=false in the
628 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
629 remote server are returned as part of bounce messages. This was not
630 happening for LMTP over a pipe (the lmtp transport), but now it is the
631 same for both kinds of LMTP.
633 PH/38 Despite being documented as not happening, Exim was rewriting addresses
634 in header lines that were in fact CNAMEs. This is no longer the case.
636 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
637 and queue runs started by the daemon processed all messages. This has
638 been fixed so that -R and -S can now usefully be given with -q<time>.
640 PH/40 Import PCRE release 6.7 (fixes some bugs).
642 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
644 PH/42 Give an error if -q is specified more than once.
646 PH/43 Renamed the variables $interface_address and $interface_port as
647 $received_ip_address and $received_port, to make it clear that these
648 values apply to message reception, and not to the outgoing interface when
649 a message is delivered. (The old names remain recognized, of course.)
651 PH/44 There was no timeout on the connect() call when using a Unix domain
652 socket in the ${readsocket expansion. There now is.
654 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
655 be meaningful with "accept".
657 SC/01 Eximstats V1.43
658 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
660 SC/02 Eximstats V1.44
661 Use a glob alias rather than an array ref in the generated
662 parser. This improves both readability and performance.
664 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
665 Collect SpamAssassin and rejection statistics.
666 Don't display local sender or destination tables unless
667 there is data to show.
668 Added average volumes into the top table text output.
670 SC/04 Eximstats V1.46
671 Collect data on the number of addresses (recipients)
672 as well as the number of messages.
674 SC/05 Eximstats V1.47
675 Added 'Message too big' to the list of mail rejection
676 reasons (thanks to Marco Gaiarin).
678 SC/06 Eximstats V1.48
679 Mainlog lines which have GMT offsets and are too short to
680 have a flag are now skipped.
682 SC/07 Eximstats V1.49 (Alain Williams)
683 Added the -emptyok flag.
685 SC/08 Eximstats V1.50
686 Fixes for obtaining the IP address from reject messages.
688 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
689 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
690 whitesspace changes from 4.64-PH/27
692 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
695 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
696 are found, allow negative numbers in numeric criteria)
698 JJ/06 exipick.20061117.2, added new $message_body_missing variable
700 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
701 to match changes made in 4.64-PH/43
703 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
705 PH/47 Put in an explicit test for a DNS lookup of an address record where the
706 "domain" is actually an IP address, and force a failure. This locks out
707 those revolvers/nameservers that support "A-for-A" lookups, in
708 contravention of the specifications.
710 PH/48 When a host name was looked up from an IP address, and the subsequent
711 forward lookup of the name timed out, the host name was left in
712 $sender_host_name, contrary to the specification.
714 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
715 restricted to single-key lookups, Exim was not diagnosing an error if
716 * or *@ was used with a query-style lookup.
718 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
720 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
721 long ago, but noone interested enough thought of it. Let's just say that
722 the "1.1" means that there are some new functions that weren't there at
723 some point in the past.
725 PH/51 Error processing for expansion failure of helo_data from an smtp
726 transport during callout processing was broken.
728 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
729 tested/used via the -bh/-bhc/-bs options.
731 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
732 bug, fixed in subsequent PCRE releases).
734 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
735 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
737 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
743 SC/01 Use a glob alias rather than an array ref in eximstats generated
744 parser. This improves both readability and performance.
746 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
747 Don't display local sender or destination tables in eximstats unless
748 there is data to show.
749 Added average volumes into the eximstats top table text output.
751 SC/03 Collect data on the number of addresses (recipients) as well
752 as the number of messages in eximstats.
754 TF/01 Correct an error in the documentation for the redirect router. Exim
755 does (usually) call initgroups() when daemonizing.
757 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
758 with consistent privilege compared to when running as a daemon.
760 TF/03 Note in the spec that $authenticated_id is not set for local
761 submissions from trusted users.
763 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
764 Thanks to Dean Brooks <dean@iglou.com> for the patch.
766 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
767 by adding some example configuration directives to the default
768 configuration file. A little bit of work is required to uncomment the
769 directives and define how usernames and passwords are checked, but
770 there is now a framework to start from.
772 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
773 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
774 without this. I don't know how relevant this is to other LDAP libraries.
776 PH/02 Add the verb name to the "unknown ACL verb" error.
778 PH/03 Magnus Holmgren's patch for filter_prepend_home.
780 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
782 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
783 directory not expanded when it should be if an expanded home directory
784 was set for the address (which is overridden by the transport).
786 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
789 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
790 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
791 because it is too late at that time, and has no effect.
793 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
794 security issue with \' (bugzilla #107). I could not use the
795 PQescapeStringConn() function, because it needs a PGconn value as one of
798 PH/08 When testing addresses using -bt, indicate those final addresses that
799 are duplicates that would not cause an additional delivery. At least one
800 person was confused, thinking that -bt output corresponded to deliveries.
801 (Suppressing duplicates isn't a good idea as you lose the information
802 about possibly different redirections that led to the duplicates.)
804 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
805 systems where poll() doesn't work, in particular OS X.
807 PH/10 Added more information to debugging output for retry time not reached.
809 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
810 operations in malware.c.
812 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
815 PH/13 If write_rejectlog was set false when logging was sent to syslog with
816 syslog_duplication set false, log lines that would normally be written
817 both the the main log and to the reject log were not written to syslog at
820 PH/14 In the default configuration, change the use of "message" in ACL warn
821 statements to "add_header".
823 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
824 not followed by a command (e.g. "seen endif").
826 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
827 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
830 PH/17 Added extra conditions to the default value of delay_warning_condition
834 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
835 { match{$h_precedence:}{(?i)bulk|list|junk} } \
836 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
839 The Auto-Submitted: and various List- headers are standardised, whereas I
840 don't think Precedence: ever was.
842 PH/18 Refactored debugging code in route_finduser() to show more information,
843 in particular, the error code if getpwnam() issues one.
845 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
846 This is apparently needed in addition to the PH/07 change above to avoid
847 any possible encoding problems.
849 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
850 but not after initializing Perl.
852 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
853 output them only if debugging. By default they are written stderr,
854 apparently, which is not desirable.
856 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
859 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
862 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
864 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
865 authenticated or an ident call has been made. Suppress the default
866 values for $authenticated_id and $authenticated_sender (but permit -oMai
867 and -oMas) when testing with -bh.
869 PH/24 Re-jigged the order of the tests in the default configuration so that the
870 tests for valid domains and recipients precede the DNS black list and CSA
871 tests, on the grounds that those ones are more expensive.
873 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
874 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
875 This bug exists in every version of Exim that I still have, right back to
878 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
879 However, an attempt to turn on case-sensitivity in a regex key by
880 including (?-i) didn't work because the subject string was already
881 lowercased, and the effects were non-intuitive. It turns out that a
882 one-line patch can be used to allow (?-i) to work as expected.
888 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
889 other effects) broke the use of negated acl sub-conditions.
891 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
894 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
895 "Deny" causes Exim to reject the incoming connection with a 554 error.
896 Unfortunately, if there is a major crisis, such as a disk failure,
897 tcp-wrappers gives "deny", whereas what one would like would be some
898 kind of temporary error. A kludge has been added to help with this.
899 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
900 554 error is used if errno is still zero or contains ENOENT (which occurs
901 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
904 PH/03 Add -lutil to the default FreeBSD LIBS setting.
906 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
907 errors. Otherwise a message that provokes a temporary error (when other
908 messages do not) can cause a whole host to time out.
910 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
911 addresses were routed directly to files or pipes from a redirect router.
912 File deliveries just didn't batch; pipe deliveries might have suffered
915 PH/06 A failure to get a lock for a hints database would erroneously always say
916 "Failed to get write lock", even when it was really a read lock.
918 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
919 of 0600. This has been changed to use the value of the lockfile_mode
920 option (which defaults to 0600).
922 PH/08 Applied small patch from the Sieve maintainer.
924 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
925 folder from quota calculations, a direct delivery into this folder messed
926 up the contents of the maildirsize file. This was because the regex was
927 used only to exclude .Trash (or whatever) when the size of the mailbox
928 was calculated. There was no check that a delivery was happening into an
929 excluded directory. This bug has been fixed by ignoring all quota
930 processing for deliveries into excluded directories.
932 PH/10 Added the maildirfolder_create_regex option to appendfile.
938 PH/01 The code for finding all the local interface addresses on a FreeBSD
939 system running IPv6 was broken. This may well have applied to all BSD
940 systems, as well as to others that have similar system calls. The broken
941 code found IPv4 interfaces correctly, but gave incorrect values for the
942 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
943 that it would not match correctly against @[] and not recognize the IPv6
946 PH/02 The ipliteral router was not recognizing addresses of the form user@
947 [ipv6:....] because it didn't know about the "ipv6:" prefix.
949 PH/03 Added disable_ipv6.
951 PH/04 Changed $reply_address to use the raw form of the headers instead of the
952 decoded form, because it is most often used to construct To: headers
953 lines in autoreplies, and the decoded form may well be syntactically
954 invalid. However, $reply_address has leading white space removed, and all
955 newlines turned into spaces so that the autoreply transport does not
958 PH/05 If group was specified without a user on a router, and no group or user
959 was specified on a transport, the group from the router was ignored.
961 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
962 for visible compile-time settings that can be used to change these
963 numbers, for those that want even more. Backwards compatibility with old
964 spool files has been maintained. However, going back to a previous Exim
965 release will lost any variables that are in spool files.
967 PH/07 Two small changes when running in the test harness: increase delay when
968 passing a TCP/IP connection to a new process, in case the original
969 process has to generate a bounce, and remove special handling of
970 127.0.0.2 (sic), which is no longer necessary.
972 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
973 be the same on different OS.
975 PH/09 Moved a debug statement in filter processing to avoid a race problem when
978 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
979 whether --show-vars was specified or not
981 JJ/02 exipick: Added support for new ACL variable spool format introduced
984 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
985 syntactically invalid From: or Reply-to: line, and a filter used this to
986 generate an autoreply, and therefore failed to obtain an address for the
987 autoreply, Exim could try to deliver to a non-existent relative file
988 name, causing unrelated and misleading errors. What now happens is that
989 it logs this as a hard delivery error, but does not attempt to create a
992 PH/11 The exinext utility has a -C option for testing purposes, but although
993 the given file was scanned by exinext itself; it wasn't being passed on
994 when Exim was called.
996 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
997 an end-of-file indication when reading a command response.
999 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1000 compiled. In many other places in Exim, IPv6 addresses are always
1001 recognized, so I have changed this. It also means that IPv4 domain
1002 literals of the form [IPV4:n.n.n.n] are now always recognized.
1004 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1005 used if the router is not running as root, for example, when verifying at
1006 ACL time, or when using -bh. The debugging output from this situation was
1007 non-existent - all you got was a failure to exec. I have made two
1010 (a) Failures to set uid/gid, the current directory, or a process leader
1011 in a subprocess such as that created by queryprogram now generate
1012 suitable debugging ouput when -d is set.
1014 (b) The queryprogram router detects when it is not running as root,
1015 outputs suitable debugging information if -d is set, and then runs
1016 the subprocess without attempting to change uid/gid.
1018 PH/15 Minor change to Makefile for building test_host (undocumented testing
1021 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1022 additional section of a DNS packet that returns MX or SRV records.
1023 Instead, it always explicitly searches for A/AAAA records. This avoids
1024 major problems that occur when a DNS server includes only records of one
1025 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1026 fixed another bug: if SRV records were looked up and the corresponding
1027 address records were *not* found in the additional section, the port
1028 values from the SRV records were lost.
1030 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1031 using the correct key (the original address) when searching the retry
1032 rules in order to find which one to use for generating the retry hint.
1034 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1035 adding the default one. Similarly, if it contains a Reply-To: header, the
1036 errors_reply_to option, if set, is not used.
1038 PH/19 When calculating a retry time, Exim used to measure the "time since
1039 failure" by looking at the "first failed" field in the retry record. Now
1040 it does not use this if it is later than than the arrival time of the
1041 message. Instead it uses the arrival time. This makes for better
1042 behaviour in cases where some deliveries succeed, thus re-setting the
1043 "first failed" field. An example is a quota failure for a huge message
1044 when small messages continue to be delivered. Without this change, the
1045 "time since failure" will always be short, possible causing more frequent
1046 delivery attempts for the huge message than are intended.
1047 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1049 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1050 $1, $2, $3) because the numerical variables can be reset during some
1051 expansion items (e.g. "match"), thereby losing the authentication data.
1053 PH/21 Make -bV show the size of off_t variables so that the test suite can
1054 decide whether to run tests for quotas > 2G.
1056 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1057 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1058 filecount value is greater than 2G or if a quota value is greater than 2G
1059 on a system where the size of off_t is not greater than 4, a panic error
1062 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1063 never match. The debug and -bh output now contains an explicit error
1064 message indicating a malformed IPv4 address or mask.
1066 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1067 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1068 PH/23 above applies.
1070 PH/25 Do not write to syslog when running in the test harness. The only
1071 occasion when this arises is a failure to open the main or panic logs
1072 (for which there is an explicit test).
1074 PH/26 Added the /no_tell option to "control=freeze".
1076 PH/27 If a host name lookup failed very early in a connection, for example, if
1077 the IP address matched host_lookup and the reverse lookup yielded a name
1078 that did not have a forward lookup, an error message of the form "no IP
1079 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1080 could be logged. Now it outputs the IP address instead of "NULL".
1082 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1083 allows the sender and the authenticated sender to be set when
1084 submitting a message from within Exim. Since child_open_exim() is
1085 documented for local_scan(), the new function should be too.
1087 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1088 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1089 results in an empty string is now treated as unset.
1091 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1093 PH/31 Added sender_verify_fail logging option.
1095 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1096 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1097 tidied the source and removed it altogether.
1099 PH/33 When a queue run was abandoned because the load average was too high, a
1100 log line was always written; now it is written only if the queue_run log
1101 selector is set. In addition, the log line for abandonment now contains
1102 information about the queue run such as the pid. This is always present
1103 in "start" and "stop" lines but was omitted from the "abandon" line.
1105 PH/34 Omit spaces between a header name and the colon in the error message that
1106 is given when verify = headers_syntax fails (if there are lots of them,
1107 the message gets confusing).
1109 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1110 names, as there are now some PTR records that contain slashes. This check
1111 is only to protect against broken name servers that fall over on strange
1112 characters, so the fact that it applies to all lookups doesn't matter.
1114 PH/36 Now that the new test suite is complete, we can remove some of the
1115 special code in Exim that was needed for the old test suite. For example,
1116 sorting DNS records because real resolvers return them in an arbitrary
1117 order. The new test suite's fake resolver always returns records in the
1120 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1121 bounces) except when queue_only is set, to avoid logging races between
1122 the different processes.
1124 PH/38 Panic-die if .include specifies a non-absolute path.
1126 PH/39 A tweak to the "H" retry rule from its user.
1128 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1129 a label. They prevented compilation on older perls.
1131 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1132 a warning to be raised on newish perls.
1134 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1135 on queue. Changes to match documented behaviour of showing count of
1136 messages matching specified criteria.
1138 PH/40 Changed the default ident timeout from 30s to 5s.
1140 PH/41 Added support for the use of login_cap features, on those BSD systems
1141 that have them, for controlling the resources used by pipe deliveries.
1143 PH/42 The content-scanning code uses fopen() to create files in which to put
1144 message data. Previously it was not paying any attention to the mode of
1145 the files. Exim runs with umask(0) because the rest of the code creates
1146 files with open(), and sets the required mode explicitly. Thus, these
1147 files were ending up world-writeable. This was not a big issue, because,
1148 being within the spool directory, they were not world-accessible. I have
1149 created a function called modefopen, which takes an additional mode
1150 argument. It sets umask(777), creates the file, chmods it to the required
1151 mode, then resets the umask. All the relevant calls to fopen() in the
1152 content scanning code have been changed to use this function.
1154 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1155 to 24 hours. This avoids potential overflow problems when processing G
1156 and H retry rules. I suspect nobody ever tinkers with this value.
1158 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1160 PH/45 When the plaintext authenticator is running as a client, the server's
1161 challenges are checked to ensure they are valid base64 strings. By
1162 default, the authentication attempt is cancelled if an invalid string is
1163 received. Setting client_ignore_invalid_base64 true ignores these errors.
1164 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1165 they are received. Thus, the responses can be made to depend on the
1166 challenges. If an invalid string is ignored, an empty string is placed in
1169 PH/46 Messages that are created by the autoreply transport now contains a
1170 References: header, in accordance with RFCs 2822 and 3834.
1172 PH/47 Added authenticated_sender_force to the smtp transport.
1174 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1176 PH/49 Installed latest patch from the Sieve maintainer.
1178 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1179 was also set, the appendfile transport was unnecessarily scanning a
1180 directory of message files (e.g. for maildir delivery) to find the count
1181 of files (along with the size), even though it did not need this
1182 information. It now does the scan only if it needs to find either the
1183 size of the count of files.
1185 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1187 PH/52 Two bugs concerned with error handling when the smtp transport is
1190 (i) Exim was not creating retry information for temporary errors given
1191 for individual recipients after the DATA command when the smtp transport
1192 was used in LMTP mode. This meant that they could be retried too
1193 frequently, and not timed out correctly.
1195 (ii) Exim was setting the flag that allows error details to be returned
1196 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1197 recipients that were returned after the DATA command.
1199 PH/53 This is related to PH/52, but is more general: for any failing address,
1200 when detailed error information was permitted to be returned to the
1201 sender, but the error was temporary, then after the final timeout, only
1202 "retry timeout exceeded" was returned. Now it returns the full error as
1203 well as "retry timeout exceeded".
1205 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1206 do this, and (what is worse) MTAs that accept it.
1208 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1209 will now be deprecated.
1211 PH/56 New os.c-cygwin from the Cygwin maintainer.
1213 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1214 formats (previously only available in exim formats via -bpr, -bpru,
1215 and -bpra. Now also available in native and exiqgrep formats)
1217 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1218 with very large, slow to parse queues
1220 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1222 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1224 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1225 responses to authentication challenges, though it was showing the
1226 challenges; (ii) I've removed the CR characters from the debug output for
1229 PH/58 Allow for the insertion of a newline as well as a space when a string
1230 is turned into more than one encoded-word during RFC 2047 encoding. The
1231 Sieve code now uses this.
1233 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1234 data_4xx, lost_connection, tls_required.
1236 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1237 message was being sent as an SMTP response.
1239 PH/61 Add -l and -k options to exicyclog.
1241 PH/62 When verifying, if an address was redirected to one new address, so that
1242 verification continued, and the new address failed or deferred after
1243 having set something in $address_data, the value of $address_data was not
1244 passed back to the ACL. This was different to the case when no
1245 redirection occurred. The value is now passed back in both cases.
1247 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1248 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1249 use login_cap.h, so on its own it isn't the distinguishing feature. The
1250 new name refers directly to the setclassresources() function.
1252 PH/65 Added configuration files for NetBSD3.
1254 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1256 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1257 is preferred over IPv4.
1259 PH/68 The bounce_return_message and bounce_return_body options were not being
1260 honoured for bounces generated during the reception of non-SMTP messages.
1261 In particular, this applied to messages rejected by the ACL. This bug has
1262 been fixed. However, if bounce_return_message is true and bounce_return_
1263 body is false, the headers that are returned for a non-SMTP message
1264 include only those that have been read before the error was detected.
1265 (In the case of an ACL rejection, they have all been read.)
1267 PH/69 The HTML version of the specification is now built in a directory called
1268 spec_html instead of spec.html, because the latter looks like a path with
1269 a MIME-type, and this confuses some software.
1271 PH/70 Catch two compiler warnings in sieve.c.
1273 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1274 function verify_get_ident() calls ip_connect() to connect a socket, but
1275 if the "connect()" function timed out, ip_connect() used to close the
1276 socket. However, verify_get_ident() also closes the socket later, and in
1277 between Exim writes to the log, which may get opened at this point. When
1278 the socket was closed in ip_connect(), the log could get the same file
1279 descriptor number as the socket. This naturally causes chaos. The fix is
1280 not to close the socket in ip_connect(); the socket should be closed by
1281 the function that creates it. There was only one place in the code where
1282 this was missing, in the iplookup router, which I don't think anybody now
1283 uses, but I've fixed it anyway.
1285 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1286 well as to direct DNS lookups. Otherwise the handling of names in host
1287 lists is inconsistent and therefore confusing.
1293 PH/01 Two changes to the default runtime configuration:
1295 (1) Move the checks for relay_from_hosts and authenticated clients from
1296 after to before the (commented out) DNS black list checks.
1298 (2) Add control=submission to the relay_from_hosts and authenticated
1299 clients checks, on the grounds that messages accepted by these
1300 statements are most likely to be submissions.
1302 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1304 (1) Generate an error if the third argument for the ${prvs expansion is
1307 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1310 (3) Reset the variables that are obtained from the first argument of
1311 ${prvscheck and used in the second argument before leaving the code,
1312 because their memory is reclaimed, so using them afterwards may do
1315 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1316 one (it's much easier than Tom thought :-).
1318 (5) Because of (4), we can now allow for the use of $prvscheck_result
1319 inside the third argument.
1321 PH/03 For some reason, the default setting of PATH when running a command from
1322 a pipe transport was just "/usr/bin". I have changed it to
1325 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1326 anything to be listed in the output from -bV.
1328 PH/05 When a filter generated an autoreply, the entire To: header line was
1329 quoted in the delivery log line, like this:
1331 => >A.N.Other <ano@some.domain> <original@ddress> ...
1333 This has been changed so that it extracts the operative address. There
1334 may be more than one such address. If so, they are comma-separated, like
1337 => >ano@some.domain,ona@other.domain <original@ddress> ...
1339 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1340 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1341 not being looked up in the rDNS to get a host name, Exim was showing the
1342 IP address twice in Received: lines, even though the IP addresses were
1343 identical. For example:
1345 Received: from [1.2.3.4] (helo=[1.2.3.4])
1347 However, if the real host name was known, it was omitting the HELO data
1348 if it matched the actual IP address. This has been tidied up so that it
1349 doesn't show the same IP address twice.
1351 PH/07 When both +timestamp and +memory debugging was on, the value given by
1352 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1353 called by the debug printing, thereby overwriting the timestamp buffer.
1354 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1356 PH/08 When the original message was included in an autoreply transport, it
1357 always said "this is a copy of the message, including all the headers",
1358 even if body_only or headers_only was set. It now gives an appropriate
1361 PH/09 Applied a patch from the Sieve maintainer which:
1363 o fixes some comments
1364 o adds the (disabled) notify extension core
1365 o adds some debug output for the result of if/elsif tests
1366 o points to the current vacation draft in the documentation
1367 and documents the missing references header update
1371 o fixes a bug in processing the envelope test (when testing
1372 multiple envelope elements, the last element determinted the
1375 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1376 Electronic Mail") by including:
1378 Auto-submitted: auto-generated
1380 in the messages that it generates (bounce messages and others, such as
1381 warnings). In the case of bounce messages for non-SMTP mesages, there was
1382 also a typo: it was using "Auto_submitted" (underscore instead of
1383 hyphen). Since every message generated by Exim is necessarily in response
1384 to another message, thes have all been changed to:
1386 Auto-Submitted: auto-replied
1388 in accordance with these statements in the RFC:
1390 The auto-replied keyword:
1392 - SHOULD be used on messages sent in direct response to another
1393 message by an automatic process,
1395 - MUST NOT be used on manually-generated messages,
1397 - MAY be used on Delivery Status Notifications (DSNs) and Message
1398 Disposition Notifications (MDNs),
1400 - MUST NOT be used on messages generated by automatic or periodic
1401 processes, except for messages which are automatic responses to
1404 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1405 to the default Received: header definition.
1407 PH/12 Added log selector acl_warn_skipped (default on).
1409 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1410 variables because (a) they are in the wrong storage pool and (b) even if
1411 they were copied, it wouldn't work properly because of the caching.
1413 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1414 checking when decoding. Apparently there are clients that generate
1415 overlong encoded strings. Why am I not surprised?
1417 PH/15 If the first argument of "${if match_address" was not empty, but did not
1418 contain an "@" character, Exim crashed. Now it writes a panic log message
1419 and treats the condition as false.
1421 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1423 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1424 extension "envelope-auth". The code is finished and in agreement with
1425 other implementations, but there is no documentation so far and in fact,
1426 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1427 not changing the active code.
1429 Print executed "if" and "elsif" statements when debugging is used. This
1430 helps a great deal to understand what a filter does.
1432 Document more things not specified clearly in RFC3028. I had all this
1433 sorted out, when out of a sudden new issues came to my mind. Oops."
1435 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1438 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1439 canonical form (as documented). However, after a host name lookup from
1440 the IP address, check_host() was doing a simple string comparison with
1441 addresses acquired from the DNS when checking that the found name did
1442 have the original IP as one of its addresses. Since any found IPv6
1443 addresses are likely to be in abbreviated form, the comparison could
1444 fail. Luckily, there already exists a function for doing the comparison
1445 by converting both addresses to binary, so now that is used instead of
1446 the text comparison.
1448 PH/20 There was another similar case to PH/19, when a complete host name was
1449 given in a host list; looking up its IP address could give an abbreviated
1450 form, whereas the current host's name might or might not be abbreviated.
1451 The same fix has been applied.
1457 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1458 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1461 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1462 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1464 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1466 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1467 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1468 Also, for "Auto-Submitted", treat anything other than "no" as
1469 non-personal, in accordance with RFC 3834. (Previously it treated
1470 anything starting "auto-" as non-personal.)
1472 TF/01 The control=submission/name=... option had a problem with syntax
1473 errors if the name included a slash character. The /name= option
1474 now slurps the rest of the string, so it can include any characters
1475 but it must come last in the list of options (after /sender_retain
1478 PH/05 Some modifications to the interface to the fake nameserver for the new
1486 TK/01 Added the "success_on_redirect" address verification option. See
1487 NewStuff for rationale and an example.
1489 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1491 PH/02 Patch to exigrep to allow it to work on syslog lines.
1493 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1494 fread() to skip over the body file's header line, because in Cygwin the
1495 header line is locked and is inaccessible.
1497 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1498 co-exist for some time) to make it clear that it is the Exim ID that is
1499 referenced, not the Message-ID: header line.
1501 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1502 string_format() function, because snprintf() does not exist on all
1505 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1506 ${stat: expansion item.
1508 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1509 protocol synchronization error", to keep the pedants happy.
1511 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1512 well as for IRIX systems, when gcc is being used. See the host.c source
1515 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1517 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1520 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1521 transport and to the smtp transport in LMTP mode.
1523 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1525 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1526 run a filter in a subprocess. This could lead to confusion in subsequent
1527 lookups in the parent process. There should also be a search_tidyup() at
1528 the end of the subprocess.
1530 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1531 only if the host matched helo_try_verify_hosts, which caused the
1532 verification to occur when the EHLO/HELO command was issued. The ACL just
1533 tested the remembered result. Now, if a previous verification attempt has
1534 not happened, "verify = helo" does it there and then.
1536 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1538 TK/03 Fix log output including CR from clamd.
1540 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1541 exist provoked a memory error which could cause a segfault.
1543 PH/15 Installed PCRE 6.2
1545 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1547 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1548 of the problem. Specifically, suggested +O2 rather than +O1 for the
1551 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1553 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1554 record, the port was not correctly logged when the outgoing_port log
1555 selector was set (it logged the transort's default port).
1557 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1558 fallback_hosts, and "hosts" in the smtp transport.
1560 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1561 host errors such as "Connection refused".
1563 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1564 authentication with radiusclient 0.4.9:
1566 - Error returned from rc_read_config was caught wrongly
1567 - Username/password not passed on to radius server due to wrong length.
1569 The presumption is that some radiusclient API changes for 4.51/PH/17
1570 were not taken care of correctly. The code is still untested by me (my
1571 Linux distribution still has 0.3.2 of radiusclient), but it was
1572 contributed by a Radius user.
1574 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1575 expanding the "port" option of the smtp transport.
1577 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1578 while reading a MIME header. Thanks to Tom Hughes for a patch.
1580 PH/24 Include config.h inside local_scan.h so that configuration settings are
1583 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1584 that in an ACL for RCPT (for example), you can examine exactly what was
1587 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1588 commands, but it was not correctly comparing the address with the actual
1589 client host address. Thus, it would show the EHLO address in Received:
1590 header lines when this was not necessary.
1592 PH/27 Added the % operator to ${eval:}.
1594 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1595 it should be ignoring failures (because with -C, for example, it has lost
1596 privilege). It wasn't ignoring creation failures other than "already
1599 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1602 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1603 because an input line was too long, either on its own, or by virtue of
1604 too many continuations, the temporary file was not being removed, and the
1605 return code was incorrect.
1607 PH/31 Missing "BOOL" in function definition in filtertest.c.
1609 PH/32 Applied Sieve patches from the maintainer.
1611 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1613 PH/33 Added "verify = not_blind".
1615 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1616 Local/Makefile (with some defaults set). These are used in built scripts
1617 such as exicyclog, but they have never been used in the exim_install
1618 script (though there are many overriding facilities there). I have
1619 arranged that the exim_install script now takes note of these two
1622 PH/35 Installed configuration files for Dragonfly.
1624 PH/36 When a locally submitted message by a trusted user did not contain a
1625 From: header, and the sender address was obtained from -f or from an SMTP
1626 MAIL command, and the trusted user did not use -F to supply a sender
1627 name, $originator_name was incorrectly used when constructing a From:
1628 header. Furthermore, $originator_name was used for submission mode
1629 messages from external hosts without From: headers in a similar way,
1630 which is clearly wrong.
1632 PH/37 Added control=suppress_local_fixups.
1634 PH/38 When log_selector = +received_sender was set, and the addition of the
1635 sender made the log line's construction buffer exactly full, or one byte
1636 less than full, an overflow happened when the terminating "\n" was
1639 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1640 when the result of a list match is failure because a DNS lookup failed.
1642 PH/40 RM_COMMAND is now used in the building process.
1644 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1645 the "build-* directories that it finds.
1647 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1648 address in a domain literal was a prefix of an interface address.
1650 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1651 when verifying a sender address, unless rewrite_headers is false.
1653 PH/44 Wrote a long comment about why errors_to addresses are verified as
1654 recipients, not senders.
1656 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1657 the ratelimit ACL was added.
1659 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1661 PH/46 Added extra information about PostgreSQL errors to the error string.
1663 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1664 avoiding the need to install special zones in a real server. This is
1665 backwards compatible; if it can't find the fake resolver, it drops back.
1666 Thus, both old and new test suites can be run.
1668 TF/02 Added util/ratelimit.pl
1670 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1671 clock is set back in time.
1673 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1674 Candler <B.Candler@pobox.com>.
1676 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1677 OK for addresses that are the result of redirections.
1679 PH/48 A number of further additions for the benefit of the new test suite,
1680 including a fake gethostbyname() that interfaces to the fake DNS resolver
1683 TF/06 The fix for widen_domains has also been applied to qualify_single and
1684 search_parents which are the other dnslookup options that can cause
1687 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1690 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1692 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1693 service" if it encountered an MX record with an empty target hostname.
1694 The message is now "an MX or SRV record indicated no SMTP service".
1696 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1697 if the DNS of the sending site is misconfigured. This is quite a
1698 common situation. This change restores the behaviour of treating a
1699 helo verification defer as a failure.
1701 PH/51 If self=fail was set on a router, the bounce message did not include the
1702 actual error message.
1708 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1710 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1711 command itself ended in error, the underlying message about the transport
1712 filter timeout was being overwritten with the pipe command error. Now the
1713 underlying error message should be appended to the second error message.
1715 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1717 PH/02 Reduce the amount of output that "make" produces by default. Full output
1718 can still be requested.
1720 PH/03 The warning log line about a condition test deferring for a "warn" verb
1721 was being output only once per connection, rather than after each
1722 occurrence (because it was using the same function as for successful
1723 "warn" verbs). This seems wrong, so I have changed it.
1725 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1726 it should not have, which might have caused a crash in the right
1727 circumstances, but probably never did.
1729 PH/04 Installed a modified version of Tony Finch's patch to make submission
1730 mode fix the return path as well as the Sender: header line, and to
1731 add a /name= option so that you can make the user's friendly name appear
1734 TF/03 Added the control = fakedefer ACL modifier.
1736 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1737 Mark Lowes for thorough testing.
1739 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1741 TK/03 Merged latest SRS patch from Miles Wilton.
1743 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1744 with the definition in sysexits.h (which is #included earlier).
1745 Fortunately, Exim does not actually use EX_OK. The code used to try to
1746 preserve the sysexits.h value, by assumimg that macro definitions were
1747 scanned for macro replacements. I have been disabused of this notion,
1748 so now the code just undefines EX_OK before #including unistd.h.
1750 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1751 in the smtp transport. When a block could not be written in a single
1752 write() function, the timeout was being re-applied to each part-write.
1753 This seems wrong - if the receiver was accepting one byte at a time it
1754 would take for ever. The timeout is now adjusted when this happens. It
1755 doesn't have to be particularly precise.
1757 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1758 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1760 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1761 without a domain if the check to <postmaster@domain> fails.
1763 SC/01 Eximstats: added -xls and the ability to specify output files
1764 (patch written by Frank Heydlauf).
1766 SC/02 Eximstats: use FileHandles for outputing results.
1768 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1770 SC/04 Eximstats: fixed display of large numbers with -nvr option
1772 SC/05 Eximstats: fixed merging of reports with empty tables.
1774 SC/06 Eximstats: added the -include_original_destination flag
1776 SC/07 Eximstats: removed tabs and trailing whitespace.
1778 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1780 TK/06 MBOX spool code: Add real "From " MBOX separator line
1781 so the .eml file is really in mbox format (even though
1782 most programs do not really care). Patch from Alex Miller.
1784 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1785 The latter is generated from $received_to and is only set if the
1786 message has one envelope recipient. SA can use these headers,
1787 obviously out-of-the-box. Patch from Alex Miller.
1789 PH/08 The ${def test on a variable was returning false if the variable's
1790 value was "0", contrary to what the specification has always said!
1791 The result should be true unless the variable is empty.
1793 PH/09 The syntax error of a character other than { following "${if
1794 def:variable_name" (after optional whitespace) was not being diagnosed.
1795 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1796 accidental colon was present, for example, could give incorrect results.
1798 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1799 result is used (not including appendfile, where other changes are about
1802 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1803 This involved changing a lot of size variables from int to off_t. It
1804 should work with maildirs and everything.
1806 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1807 spamd dying while we are connected to it.
1809 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1812 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1813 function for BDB 4.3.
1815 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1817 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1818 the sources for PCRE have been re-arranged and more files are now
1821 PH/15 The code I had for printing potentially long long variables in PH/11
1822 above was not the best (it lost precision). The length of off_t variables
1823 is now inspected at build time, and an appropriate printing format (%ld
1824 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1825 to be "long long int" or "long int". This is needed for the internal
1826 formatting function string_vformat().
1828 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1829 the configuration file to be ":syslog", then the script "guesses" where
1830 the logs files are, rather than using the compiled in default. In our
1831 case the guess is not the same as the compiled default, so the script
1832 suddenly stopped working when I started to use syslog. The patch checks
1833 to see if log_file_path is "". If so, it attempts to read it from exim
1834 with no configuration file to get the compiled in version, before it
1835 falls back to the previous guessing code."
1837 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1838 implementing BATV in an Exim configuration. See NewStuff for the gory
1841 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1842 Makefile that are specific to HP-UX.
1844 PH/18 If the "use_postmaster" option was set for a recipient callout together
1845 with the "random" option, the postmaster address was used as the MAIL
1846 FROM address for the random test, but not for the subsequent recipient
1847 test. It is now used for both.
1849 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1850 patch removes a few documentation additions to RFC 3028, because the
1851 latest draft now contains them. It adds the new en;ascii-case comparator
1852 and a new error check for 8bit text in MIME parts. Comparator and
1853 require names are now matched exactly. I enabled the subaddress
1854 extension, but it is not well tested yet (read: it works for me)."
1856 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1857 rework some of the code of TK/09 above to avoid the hardwired use of
1858 "%lld" and "long long". Replaced the call to snprintf() with a call to
1861 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1862 records point to non-existent hosts", "retry timeout exceeded", and
1863 "retry time not reached for any host after a long failure period".
1865 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1866 experimental DomainKeys support:
1868 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1869 (2) On an error such as an illegally used "control", the wrong name for
1870 the control was given.
1872 These problems did NOT occur unless DomainKeys support was compiled.
1874 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1876 PH/24 Added ${if match_ip condition.
1878 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1879 databases so that it will be absolutely obvious if a crash occurs in the
1880 DB library. This is a regular occurrence (often caused by mis-matched
1883 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1884 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1885 picked up on a user's system that detects such things. There doesn't seem
1886 to be a gcc warning option for this - only an attribute that has to be
1887 put on the function's prototype. It seems that in Fedora Core 4 they have
1888 set this on a number of new functions. No doubt there will be more in due
1891 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1892 specify a transport. However, if an address that was verified by such a
1893 router was the subject of a callout, Exim crashed because it tried to
1894 read the rcpt_include_affixes from the non-existent transport. Now it
1895 just assumes that the setting of that option is false. This bug was
1896 introduced by 4.51/PH/31.
1898 PH/28 Changed -d+all to exclude +memory, because that information is very
1899 rarely of interest, but it makes the output a lot bigger. People tend to
1900 do -d+all out of habit.
1902 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1903 code in os-type was giving problems when libc.so lives in lib64, like on
1906 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1907 aren't the modern standard, and it seems that some systems' include files
1908 don't always have them. Exim was already checking for some of the newer
1909 ones like T_AAAA, and defining it itself. I've added checks for all the
1910 record types that Exim uses.
1912 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1913 not automatically generating a new one, as it is supposed to. This
1914 prevented TLS from working. If the file did exist, but contained invalid
1915 data, a new version was generated, as expected. It was only the case of a
1916 non-existent file that was broken.
1918 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1919 with a change in libdomainkeys > 0.64.
1921 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1922 from DNS. If the selector record carries the flag, it now has
1923 precedence over the domain-wide flag.
1925 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1927 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1928 the use of a port name that isn't defined in /etc/services) occurred, the
1929 message was deferred as in a normal delivery, and thus remained on the
1930 spool, instead of being failed because of the mua_wrapper setting. This
1931 is now fixed, and I tidied up some of the mua_wrapper messages at the
1934 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1935 the messages in a hash of arrays rather than using individual hashes.
1936 This is a bit cleaner and results in dramatic memory savings, albeit
1937 at a slight CPU cost.
1939 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1940 as requested by Marc Sherman.
1942 SC/10 Eximstats: added histograms for user specified patterns as requested
1945 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1947 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1948 fopen() in the content-scanning modules that did not already have it.
1954 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1955 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1957 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1959 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1961 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1962 to test Sieve filters that use "vacation".
1964 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1965 that changes the way the GnuTLS parameters are stored in the cache file.
1966 The new format can be generated externally. For backward compatibility,
1967 if the data in the cache doesn't make sense, Exim assumes it has read an
1968 old-format file, and it generates new data and writes a new file. This
1969 means that you can't go back to an older release without removing the
1972 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1973 work if there are any delivery delays because "one_time" forces the
1974 parent to be marked "delivered", so its unseen clone is never tried
1975 again. For this reason, Exim now forbids the simultaneous setting of
1978 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1979 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1980 entry if you want to know the details. The fix, however, neglected to
1981 consider the case when local delivery batching is involved. The test for
1982 "previously delivered" was not happening when checking to see if an
1983 address could be batched with a previous (undelivered) one; under
1984 certain circumstances this could lead to multiple deliveries to the same
1987 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1988 in its include files, and this causes problems building Exim.
1990 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1991 header_syntax) but Exim was just ignoring anything given after a slash.
1992 In particular, this caused confusion with an attempt to use "verify =
1993 reverse_host_lookup/defer_ok". An error is now given when options are
1994 supplied for verify items that do not have them. (Maybe reverse_host_
1995 lookup should have a defer_ok option, but that's a different point.)
1997 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1998 defined by RFC 821) to 2048, because there were problems with some AUTH
1999 commands, and RFC 1869 says the size should be increased for extended
2000 SMTP commands that take arguments.
2002 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2005 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2006 "unknown" error; now it says that the functionality isn't in the binary.
2008 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2009 an address' error message when a string expansion fails (syntax or
2010 whatever). Otherwise the password may appear in the log. Following change
2011 PH/42 below, there is no longer a chance of it appearing in a bounce
2014 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2016 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2017 of its addresses was ever tried. (Bugzilla bug #2.)
2019 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2020 the result incorrectly in the debug output. (It correctly added a newline
2021 to what was transported.)
2023 TF/01 Added $received_time.
2025 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2026 commented out examples of how to interface to a virus scanner and to
2027 SpamAssassin. Also added commented examples of av_scanner and
2028 spamd_address settings.
2030 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2031 and controls are allowed in which ACLs. There were a couple of minor
2032 errors. Some of the entries in the conditions table (which is a table of
2033 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2034 negation of where the condition IS allowed.
2036 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2038 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2039 header file does not have a version number, so I've had to invent a new
2040 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2041 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2042 radiusclient), but it was contributed by a Radius user.
2044 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2045 files or directories, for OpenSSL.
2047 PH/19 When an Exim process that is running as root has to create an Exim log
2048 file, it does so in a subprocess that runs as exim:exim so as to get the
2049 ownership right at creation (otherwise, other Exim processes might see
2050 the file with the wrong ownership). There was no test for failure of this
2051 fork() call, which would lead to the process getting stuck as it waited
2052 for a non-existent subprocess. Forks do occasionally fail when resources
2053 run out. I reviewed all the other calls to fork(); they all seem to check
2056 PH/20 When checking for unexpected SMTP input at connect time (before writing
2057 the banner), Exim was not dealing correctly with a non-positive return
2058 from the read() function. If the client had disconnected by this time,
2059 the result was a log entry for a synchronization error with an empty
2060 string after "input=" when read() returned zero. If read() returned -1
2061 (an event I could not check), uninitialized data bytes were printed.
2062 There were reports of junk text (parts of files, etc) appearing after
2065 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2067 PH/22 Added support for macro redefinition, and (re)definition in between
2068 driver and ACL definitions.
2070 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2071 forgetting to use the resulting value; it was using the unexpanded value.
2073 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2074 hadn't been configured. The fix is from Juergen Kreileder, who
2075 understands it better than I do:
2077 "Here's what I see happening with three configured cyrus_sasl
2078 authenticators configured (plain, login, cram-md5):
2080 On startup auth_cyrus_sasl_init() gets called for each of these.
2081 This means three calls to sasl_listmech() without a specified mech_list.
2082 => SASL tests which mechs of all available mechs actually work
2083 => three warnings about OTP not working
2084 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2086 With the patch, sasl_listmech() also gets called three times. But now
2087 SASL's mech_list option is set to the server_mech specified in the the
2088 authenticator. Or in other words, the answer from sasl_listmech()
2089 gets limited to just the mech you're testing for (which is different
2091 => the return list contains just 'plain' or 'login', 'cram-md5' or
2092 nothing depending on the value of ob->server_mech.
2094 I've just tested the patch: Authentication still works fine,
2095 unavailable mechs specified in the exim configuration are still
2096 caught, and the auth.log warnings about OTP are gone."
2098 PH/25 When debugging is enabled, the contents of the command line are added
2099 to the debugging output, even when log_selector=+arguments is not
2102 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2103 answer is "GNU", and only if the return is "GNU/something" is the answer
2106 PH/27 $acl_verify_message is now set immediately after the failure of a
2107 verification in an ACL, and so is available in subsequent modifiers. In
2108 particular, the message can be preserved by coding like this:
2110 warn !verify = sender
2111 set acl_m0 = $acl_verify_message
2113 Previously, $acl_verify_message was set only while expanding "message"
2114 and "log_message" when a very denied access.
2116 PH/28 Modified OS/os.c-Linux with
2118 -#ifndef OS_LOAD_AVERAGE
2119 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2121 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2122 nomenclature these days.)
2124 PH/29 Installed patch from the Sieve maintainer that adds the options
2125 sieve_useraddress and sieve_subaddress to the redirect router.
2127 PH/30 In these circumstances:
2128 . Two addresses routed to the same list of hosts;
2129 . First host does not offer TLS;
2130 . First host accepts first address;
2131 . First host gives temporary error to second address;
2132 . Second host offers TLS and a TLS session is established;
2133 . Second host accepts second address.
2134 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2135 and peerdn, if requested) that were in fact used only for the second
2138 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2139 attention to any local part prefix or suffix that was matched by the
2140 router that accepted the address. It now behaves in the same way as it
2141 does for delivery: the affixes are removed from the local part unless
2142 rcpt_include_affixes is set on the transport.
2144 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2145 timeout during the DATA phase of an incoming message.
2147 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2148 applied a patch sanctioned by the Sieve maintainer.
2150 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2151 the uid or gid is negative. A case of a negative gid caused this to be
2152 noticed. The fix allows for either to be negative.
2154 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2155 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2158 PH/36 The change PH/12 above was broken. Fixed it.
2160 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2161 the grounds that routing the same address twice would always produce the
2162 same answer. This might have been true once, but it is certainly no
2163 longer true now. Routing a child address may depend on the previous
2164 routing that produced that child. Some complicated redirection strategies
2165 went wrong when messages had multiple recipients, and made Exim's
2166 behaviour dependent on the order in which the addresses were given.
2168 I have moved the duplicate checking until after the routing is complete.
2169 Exim scans the addresses that are assigned to local and remote
2170 transports, and removes any duplicates. This means that more work will be
2171 done, as duplicates will always all be routed, but duplicates are
2172 presumably rare, so I don't expect this is of any significance.
2174 For deliveries to pipes, files, and autoreplies, the duplicate checking
2175 still happens during the routing process, since they are not going to be
2178 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2179 It corrects a timeout issue with spamd. This is Ian's comment: "The
2180 background is that sometimes spamd either never reads data from a
2181 connection it has accepted, or it never writes response data. The exiscan
2182 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2183 blindly assumes that writes won't block so it may never time out."
2185 PH/39 Allow G after quota size as well as K and M.
2187 PH/40 The value set for $authenticated_id in an authenticator may not contain
2188 binary zeroes or newlines because the value is written to log lines and
2189 to spool files. There was no check on this. Now the value is run through
2190 the string_printing() function so that such characters are converted to
2191 printable escape sequences.
2193 PH/41 $message_linecount is a new variable that contains the total number of
2194 lines in the message. Compare $body_linecount, which is the count for the
2197 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2198 bounce and delay warning messages, except in certain special cases, which
2201 (a) An SMTP error message from a remote host;
2202 (b) A message specified in a :fail: redirection;
2203 (c) A message specified in a "fail" command in a system filter;
2204 (d) A message specified in a FAIL return from the queryprogram router;
2205 (e) A message specified by the cannot_route_message router option.
2207 In these cases only, Exim does include the error details in bounce and
2208 warning messages. There are also a few cases where bland messages such
2209 as "unrouteable address" or "local delivery error" are given.
2211 PH/43 $value is now also set for the "else" part of a ${run expansion.
2213 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2214 being worked on, but at least Exim now implements the latest version to
2217 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2218 process to complete was treated as a delivery failure, a timeout while
2219 writing the message to the pipe was logged, but erroneously treated as a
2220 successful delivery. Such timeouts include transport filter timeouts. For
2221 consistency with the overall process timeout, these timeouts are now
2222 treated as errors, giving rise to delivery failures by default. However,
2223 there is now a new Boolean option for the pipe transport called
2224 timeout_defer, which, if set TRUE, converts the failures into defers for
2225 both kinds of timeout. A transport filter timeout is now identified in
2228 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2229 systems where "make" and "gmake" are different, calling "gmake" at top
2230 level broke things. I've arranged for the value of $(MAKE) to be passed
2231 from the Makefile to this script so that it can call the same version of
2235 A note about Exim versions 4.44 and 4.50
2236 ----------------------------------------
2238 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2239 changes of various kinds. As a consequence, a big documentation update was
2240 needed. This delayed the release for rather longer than seemed good, especially
2241 in the light of a couple of (minor) security issues. Therefore, the changes
2242 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2243 release. So 4.44 and 4.50 are in effect two different branches that both start
2246 I have left the 4.50 change log unchanged; it contains all the changes since
2247 4.43. The change log for 4.44 is below; many of its items are identical to
2248 those for 4.50. This seems to be the most sensible way to preserve the
2249 historical information.
2255 1. Minor wording change to the doc/README.SIEVE file.
2257 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2258 computation of the current number of files was incorrect.
2260 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2261 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2264 4. Give more explanation in the error message when the command for a transport
2265 filter fails to execute.
2267 5. There are several places where Exim runs a non-Exim command in a
2268 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2269 was being done only for the command run by the queryprogram router. It is
2270 now done for all such subprocesses. The other cases are: ${run, transport
2271 filters, and the commands run by the lmtp and pipe transports.
2273 6. Added CONFIGURE_GROUP build-time option.
2275 7. Some older OS have a limit of 256 on the maximum number of file
2276 descriptors. Exim was using setrlimit() to set 1000 as a large value
2277 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2278 systems. I've change it so that if it can't get 1000, it tries for 256.
2280 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2281 was an oversight, and furthermore, ever since the addition of extra
2282 controls (e.g. 4.43/32), the checks on when to allow different forms of
2283 "control" were broken. There should now be diagnostics for all cases when a
2284 control that does not make sense is encountered.
2286 9. Added the /retain_sender option to "control=submission".
2288 10. $recipients is now available in the predata ACL (oversight).
2290 11. Tidy the search cache before the fork to do a delivery from a message
2291 received from the command line. Otherwise the child will trigger a lookup
2292 failure and thereby defer the delivery if it tries to use (for example) a
2293 cached ldap connection that the parent has called unbind on.
2295 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2296 of $address_data from the recipient verification was clobbered by the
2297 sender verification.
2299 13. The value of address_data from a sender verification is now available in
2300 $sender_address_data in subsequent conditions in the ACL statement.
2302 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2304 15. Added a new option "connect=<time>" to callout options, to set a different
2307 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2308 was its contents. (It was OK if the option was not defined at all.)
2310 17. A "Completed" log line is now written for messages that are removed from
2311 the spool by the -Mrm option.
2313 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2314 information about exactly what failed.
2316 19. Added -dd to debug only the daemon process.
2318 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2319 handles timeouts, both on the server side and network timeouts. Renamed the
2320 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2322 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2323 It is now set to "smtps".
2325 22. $host_address is now set to the target address during the checking of
2326 ignore_target_hosts.
2328 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2329 being passed; this would have caused $sender_host_name to have been used if
2330 matching the list had actually called for a host name (not very likely,
2331 since this list is usually IP addresses). A host name is now passed as
2334 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2335 code that specifies a non-restarting handler (typically sigaction() in
2336 modern systems) in an attempt to fix a rare and obscure crash bug.
2338 25. Narrowed the window for a race in the daemon that could cause it to ignore
2339 SIGCHLD signals. This is not a major problem, because they are used only to
2340 wake it up if nothing else does.
2342 26. A malformed maildirsize file could cause Exim to calculate negative values
2343 for the mailbox size or file count. Odd effects could occur as a result.
2344 The maildirsize information is now recalculated if the size or filecount
2347 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2348 support for a long time. Removed HAVE_SYS_VFS_H.
2350 28. Installed the latest version of exipick from John Jetmore.
2352 29. In an address list, if the pattern was not a regular expression, an empty
2353 subject address (from a bounce message) matched only if the pattern was an
2354 empty string. Non-empty patterns were not even tested. This was the wrong
2355 because it is perfectly reasonable to use an empty address as part of a
2356 database query. An empty address is now tested by patterns that are
2357 lookups. However, all the other forms of pattern expect the subject to
2358 contain a local part and a domain, and therefore, for them, an empty
2359 address still always fails if the pattern is not itself empty.
2361 30. Exim went into a mad DNS loop when attempting to do a callout where the
2362 host was specified on an smtp transport, and looking it up yielded more
2363 than one IP address.
2365 31. Re-factored the code for checking spool and log partition space into a
2366 function that finds that data and another that does the check. The former
2367 is then used to implement four new variables: $spool_space, $log_space,
2368 $spool_inodes, and $log_inodes.
2370 32. The RFC2047 encoding function was originally intended for short strings
2371 such as real names; it was not keeping to the 75-character limit for
2372 encoded words that the RFC imposes. It now respects the limit, and
2373 generates multiple encoded words if necessary. To be on the safe side, I
2374 have increased the buffer size for the ${rfc2047: expansion operator from
2377 33. It is now permitted to omit both strings after an "if" condition; if the
2378 condition is true, the result is "true". As before, when the second string
2379 is omitted, a false condition yields an empty string. This makes it less
2380 cumbersome to write custom ACL and router conditions.
2382 34. Failure to deliver a bounce message always caused it to be frozen, even if
2383 there was an errors_to setting on the router. The errors_to setting is now
2386 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2387 canonical form (fully expanded) before being placed in
2388 $sender_host_address.
2390 36. The table in the code that translates DNS record types into text (T_A to
2391 "A" for instance) was missing entries for NS and CNAME. It is just possible
2392 that this could have caused confusion if both these types were looked up
2393 for the same domain, because the text type is used as part of Exim's
2394 per-process caching. But the chance of anyone hitting this buglet seems
2397 37. The dnsdb lookup has been extended in a number of ways.
2399 (1) There is a new type, "zns", which walks up the domain tree until it
2400 finds some nameserver records. It should be used with care.
2402 (2) There is a new type, "mxh", which is like "mx" except that it returns
2403 just the host names, not the priorities.
2405 (3) It is now possible to give a list of domains (or IP addresses) to be
2406 looked up. The behaviour when one of the lookups defers can be
2407 controlled by a keyword.
2409 (4) It is now possible to specify the separator character for use when
2410 multiple records are returned.
2412 38. The dnslists ACL condition has been extended: it is now possible to supply
2413 a list of IP addresses and/or domains to be looked up in a particular DNS
2416 39. Added log_selector=+queue_time_overall.
2418 40. When running the queue in the test harness, wait just a tad after forking a
2419 delivery process, to get repeatability of debugging output.
2421 41. Include certificate and key file names in error message when GnuTLS fails
2422 to set them up, because the GnuTLS error message doesn't include the name
2423 of the failing file when there is a problem reading it.
2425 42. Allow both -bf and -bF in the same test run.
2427 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2429 44. The "Exiscan patch" is now merged into the mainline Exim source.
2431 45. Sometimes the final signoff response after QUIT could fail to get
2432 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2433 < 0 before doing a fflush(). This bug looks as though it goes back to the
2434 introduction of TLS in release 3.20, but "sometimes" must have been rare
2435 because the tests only now provoked it.
2437 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2438 (this can affect the format of dates).
2440 47. exim_tidydb, when checking for the continued existence of a message for
2441 which it has found a message-specific retry record, was not finding
2442 messages that were in split spool directories. Consequently, it was
2443 deleting retry records that should have stayed in existence.
2445 48. Steve fixed some bugs in eximstats.
2447 49. The SPA authentication driver was not abandoning authentication and moving
2448 on to the next authenticator when an expansion was forced to fail,
2449 contradicting the general specification for all authenticators. Instead it
2450 was generating a temporary error. It now behaves as specified.
2452 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2453 (the order specifies the preference for clients). The order is now AES256,
2454 AES128, 3DES, ARCFOUR128.
2456 51. Small patch to Sieve code - explicitly set From: when generating an
2459 52. Exim crashed if a remote delivery caused a very long error message to be
2460 recorded - for instance if somebody sent an entire SpamAssassin report back
2461 as a large number of 550 error lines. This bug was coincidentally fixed by
2462 increasing the size of one of Exim's internal buffers (big_buffer) that
2463 happened as part of the Exiscan merge. However, to be on the safe side, I
2464 have made the code more robust (and fixed the comments that describe what
2467 53. Now that there can be additional text after "Completed" in log lines (if
2468 the queue_time_overall log selector is set), a one-byte patch to exigrep
2469 was needed to allow it to recognize "Completed" as not the last thing in
2472 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2473 patch that reportedly fixes this has been added. I am not expert enough to
2474 create a test for it. This is what the patch creator wrote:
2476 "I found a little strange behaviour of ldap code when working with
2477 Windows 2003 AD Domain, where users was placed in more than one
2478 Organization Units. When I tried to give exim partial DN, the exit code
2479 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2480 But simultaneously result of request was absolutely normal ldap result,
2481 so I produce this patch..."
2483 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2484 REFERENCE, so I have modified the code to exclude the patch when that macro
2487 55. Some experimental protocols are using DNS PTR records for new purposes. The
2488 keys for these records are domain names, not reversed IP addresses. The
2489 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2490 leaves it alone. Component reversal etc. now happens only for IP addresses.
2493 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2495 57. Double the size of the debug message buffer (to 2048) so that more of very
2496 long debug lines gets shown.
2498 58. The exicyclog utility now does better if the number of log files to keep
2499 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2501 59. Two changes related to the smtp_active_hostname option:
2503 (1) $smtp_active_hostname is now available as a variable.
2504 (2) The default for smtp_banner uses $smtp_active_hostname instead
2505 of $primary_hostname.
2507 60. The host_aton() function is supposed to be passed a string that is known
2508 to be a valid IP address. However, in the case of IPv6 addresses, it was
2509 not checking this. This is a hostage to fortune. Exim now panics and dies
2510 if the condition is not met. A case was found where this could be provoked
2511 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2512 components; fortuitously, this particular loophole had already been fixed
2513 by change 4.50/55 above.
2515 If there are any other similar loopholes, the new check in host_aton()
2516 itself should stop them being exploited. The report I received stated that
2517 data on the command line could provoke the exploit when Exim was running as
2518 exim, but did not say which command line option was involved. All I could
2519 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2520 running as the user.
2523 61. There was a buffer overflow vulnerability in the SPA authentication code
2524 (which came originally from the Samba project). I have added a test to the
2525 spa_base64_to_bits() function which I hope fixes it.
2528 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2529 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2531 63. The daemon start-up calls getloadavg() while still root for those OS that
2532 need the first call to be done as root, but it missed one case: when
2533 deliver_queue_load_max is set with deliver_drop_privilege. This is
2534 necessary for the benefit of the queue runner, because there is no re-exec
2535 when deliver_drop_privilege is set.
2537 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2538 This has been fixed.
2540 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2541 was in use, was not putting the data itself into the right store pool;
2542 consequently, it could be overwritten for a subsequent message in the same
2543 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2546 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2548 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2549 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2550 the function were treating the return as a boolean value, which happened to
2551 work because 0=false and not-0=true, but is not correct code.
2553 68. The host_aton() function was not handling scoped IPv6 addresses (those
2554 with, for example, "%eth0" on the end) correctly.
2556 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2557 negated items (that is, ~something) in unsigned ints. Some compilers
2558 apparently mutter when there is no cast.
2560 70. If an address verification called from an ACL failed, and did not produce a
2561 user-specific message (i.e. there was only a "system" message), nothing was
2562 put in $acl_verify_message. In this situation, it now puts the system
2565 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2566 session; change 4.31/43 added the unwanted input to the log line - except
2567 that it did not do this in the start of session case. It now does.
2569 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2570 This is wrong and can cause the other end to generate a synchronization
2571 error if it is another Exim or anything else that does the synchronization
2572 check. A QUIT command is no longer sent after a timeout.
2574 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2575 during host lookups.
2577 74. The defer_ok option of callout verification was not working if it was used
2578 when verifying addresses in header lines, that is, for this case:
2580 verify = header_sender/callout=defer_ok
2582 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2583 those file descriptors could be used for SMTP connections. If anything
2584 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2585 could be sent to the SMTP client, causing chaos. The daemon now opens
2586 stdin, stdout, and stderr to /dev/null when it puts itself into the
2589 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2590 log by default. The user can override this with suitable Perl magic.
2592 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2593 the log message when discard triggers, was not working for the DATA ACL or
2594 for the non-SMTP ACL.
2596 78. Error message wording change in sieve.c.
2598 79. If smtp_accept_max_per_host was set, the number of connections could be
2599 restricted to fewer than expected, because the daemon was trying to set up
2600 a new connection before checking whether the processes handling previous
2601 connections had finished. The check for completed processes is now done
2602 earlier. On busy systems, this bug wouldn't be noticed because something
2603 else would have woken the daemon, and it would have reaped the completed
2606 80. If a message was submitted locally by a user whose login name contained one
2607 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2608 It caused a spool format error. I have fixed the spool reading code. A
2609 related problem was that the "from" clause in the Received: line became
2610 illegal because of the space(s). It is now covered by ${quote_local_part.
2612 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2615 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2617 83. Patch from Sieve maintainer to fix unterminated string problem in
2618 "vacation" handling.
2620 84. Some minor changes to the Linux configuration files to help with other
2621 OS variants using glibc.
2623 85. One more patch for Sieve to update vacation handling to latest spec.
2626 ----------------------------------------------------
2627 See the note above about the 4.44 and 4.50 releases.
2628 ----------------------------------------------------
2634 1. Change 4.43/35 introduced a bug that caused file counts to be
2635 incorrectly computed when quota_filecount was set in an appendfile
2638 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2639 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2642 3. Give more explanation in the error message when the command for a transport
2643 filter fails to execute.
2645 4. There are several places where Exim runs a non-Exim command in a
2646 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2647 was being done only for the command run by the queryprogram router. It is
2648 now done for all such subprocesses. The other cases are: ${run, transport
2649 filters, and the commands run by the lmtp and pipe transports.
2651 5. Some older OS have a limit of 256 on the maximum number of file
2652 descriptors. Exim was using setrlimit() to set 1000 as a large value
2653 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2654 systems. I've change it so that if it can't get 1000, it tries for 256.
2656 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2657 was an oversight, and furthermore, ever since the addition of extra
2658 controls (e.g. 4.43/32), the checks on when to allow different forms of
2659 "control" were broken. There should now be diagnostics for all cases when a
2660 control that does not make sense is encountered.
2662 7. $recipients is now available in the predata ACL (oversight).
2664 8. Tidy the search cache before the fork to do a delivery from a message
2665 received from the command line. Otherwise the child will trigger a lookup
2666 failure and thereby defer the delivery if it tries to use (for example) a
2667 cached ldap connection that the parent has called unbind on.
2669 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2670 of $address_data from the recipient verification was clobbered by the
2671 sender verification.
2673 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2674 was its contents. (It was OK if the option was not defined at all.)
2676 11. A "Completed" log line is now written for messages that are removed from
2677 the spool by the -Mrm option.
2679 12. $host_address is now set to the target address during the checking of
2680 ignore_target_hosts.
2682 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2683 being passed; this would have caused $sender_host_name to have been used if
2684 matching the list had actually called for a host name (not very likely,
2685 since this list is usually IP addresses). A host name is now passed as
2688 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2689 code that specifies a non-restarting handler (typically sigaction() in
2690 modern systems) in an attempt to fix a rare and obscure crash bug.
2692 15. Narrowed the window for a race in the daemon that could cause it to ignore
2693 SIGCHLD signals. This is not a major problem, because they are used only to
2694 wake it up if nothing else does.
2696 16. A malformed maildirsize file could cause Exim to calculate negative values
2697 for the mailbox size or file count. Odd effects could occur as a result.
2698 The maildirsize information is now recalculated if the size or filecount
2701 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2702 support for a long time. Removed HAVE_SYS_VFS_H.
2704 18. Updated exipick to current release from John Jetmore.
2706 19. Allow an empty sender to be matched against a lookup in an address list.
2707 Previously the only cases considered were a regular expression, or an
2710 20. Exim went into a mad DNS lookup loop when doing a callout where the
2711 host was specified on the transport, if the DNS lookup yielded more than
2714 21. The RFC2047 encoding function was originally intended for short strings
2715 such as real names; it was not keeping to the 75-character limit for
2716 encoded words that the RFC imposes. It now respects the limit, and
2717 generates multiple encoded words if necessary. To be on the safe side, I
2718 have increased the buffer size for the ${rfc2047: expansion operator from
2721 22. Failure to deliver a bounce message always caused it to be frozen, even if
2722 there was an errors_to setting on the router. The errors_to setting is now
2725 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2726 canonical form (fully expanded) before being placed in
2727 $sender_host_address.
2729 24. Updated eximstats to version 1.33
2731 25. Include certificate and key file names in error message when GnuTLS fails
2732 to set them up, because the GnuTLS error message doesn't include the name
2733 of the failing file when there is a problem reading it.
2735 26. Expand error message when OpenSSL has problems setting up cert/key files.
2738 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2739 (this can affect the format of dates).
2741 28. exim_tidydb, when checking for the continued existence of a message for
2742 which it has found a message-specific retry record, was not finding
2743 messages that were in split spool directories. Consequently, it was
2744 deleting retry records that should have stayed in existence.
2746 29. eximstats updated to version 1.35
2747 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2748 1.35 - bugfix such that pie charts by volume are generated correctly
2750 30. The SPA authentication driver was not abandoning authentication and moving
2751 on to the next authenticator when an expansion was forced to fail,
2752 contradicting the general specification for all authenticators. Instead it
2753 was generating a temporary error. It now behaves as specified.
2755 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2756 (the order specifies the preference for clients). The order is now AES256,
2757 AES128, 3DES, ARCFOUR128.
2759 31. Small patch to Sieve code - explicitly set From: when generating an
2762 32. Exim crashed if a remote delivery caused a very long error message to be
2763 recorded - for instance if somebody sent an entire SpamAssassin report back
2764 as a large number of 550 error lines. This bug was coincidentally fixed by
2765 increasing the size of one of Exim's internal buffers (big_buffer) that
2766 happened as part of the Exiscan merge. However, to be on the safe side, I
2767 have made the code more robust (and fixed the comments that describe what
2770 33. Some experimental protocols are using DNS PTR records for new purposes. The
2771 keys for these records are domain names, not reversed IP addresses. The
2772 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2773 leaves it alone. Component reversal etc. now happens only for IP addresses.
2776 34. The host_aton() function is supposed to be passed a string that is known
2777 to be a valid IP address. However, in the case of IPv6 addresses, it was
2778 not checking this. This is a hostage to fortune. Exim now panics and dies
2779 if the condition is not met. A case was found where this could be provoked
2780 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2781 components; fortuitously, this particular loophole had already been fixed
2782 by change 4.50/55 or 4.44/33 above.
2784 If there are any other similar loopholes, the new check in host_aton()
2785 itself should stop them being exploited. The report I received stated that
2786 data on the command line could provoke the exploit when Exim was running as
2787 exim, but did not say which command line option was involved. All I could
2788 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2789 running as the user.
2792 35. There was a buffer overflow vulnerability in the SPA authentication code
2793 (which came originally from the Samba project). I have added a test to the
2794 spa_base64_to_bits() function which I hope fixes it.
2797 36. The daemon start-up calls getloadavg() while still root for those OS that
2798 need the first call to be done as root, but it missed one case: when
2799 deliver_queue_load_max is set with deliver_drop_privilege. This is
2800 necessary for the benefit of the queue runner, because there is no re-exec
2801 when deliver_drop_privilege is set.
2803 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2804 was in use, was not putting the data itself into the right store pool;
2805 consequently, it could be overwritten for a subsequent message in the same
2806 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2809 38. Sometimes the final signoff response after QUIT could fail to get
2810 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2811 < 0 before doing a fflush(). This bug looks as though it goes back to the
2812 introduction of TLS in release 3.20, but "sometimes" must have been rare
2813 because the tests only now provoked it.
2819 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2820 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2821 Now it returns an integer. A number of calls were still expecting a T/F
2822 return. Fortuitously, in all cases, the tests worked in OK situations,
2823 which is the norm. However, things would have gone wrong on any write
2824 failures on the smtp file descriptor. This function is used when sending
2825 messages over SMTP and also when doing verify callouts.
2827 2. When Exim is called to do synchronous delivery of a locally submitted
2828 message (the -odf or -odi options), it no longer closes stderr before doing
2831 3. Implemented the mua_wrapper option.
2833 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2835 5. Implemented the functions header_remove(), header_testname(),
2836 header_add_at_position(), and receive_remove_recipient(), and exported them
2839 6. If an ACL "warn" statement specified the addition of headers, Exim already
2840 inserted X-ACL-Warn: at the start if there was no header name. However, it
2841 was not making this test for the second and subsequent header lines if
2842 there were newlines in the string. This meant that an invalid header could
2843 be inserted if Exim was badly configured.
2845 7. Allow an ACL "warn" statement to add header lines at the start or after all
2846 the Received: headers, as well as at the end.
2848 8. Added the rcpt_4xx retry error code.
2850 9. Added postmaster_mailfrom=xxx to callout verification option.
2852 10. Added mailfrom=xxxx to the callout verification option, for verify=
2855 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2856 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2858 12. Inserted some casts to stop certain compilers warning when using pointer
2859 differences as field lengths or precisions in printf-type calls (mostly
2860 affecting debugging statements).
2862 13. Added optional readline() support for -be (dynamically loaded).
2864 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2865 same clock tick as a message's arrival, so that its received time was the
2866 same as the "first fail" time on the retry record, and that message
2867 remained on the queue past the ultimate address timeout, every queue runner
2868 would try a delivery (because it was past the ultimate address timeout) but
2869 after another failure, the ultimate address timeout, which should have then
2870 bounced the address, did not kick in. This was a "< instead of <=" error;
2871 in most cases the first failure would have been in the next clock tick
2872 after the received time, and all would be well.
2874 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2875 being recognized when the domain list was tested by the match_domain
2876 condition in an expansion string.
2878 16. Added the ${str2b64: operator.
2880 17. Exim was always calling setrlimit() to set a large limit for the number of
2881 processes, without checking whether the existing limit was already
2882 adequate. (It did check for the limit on file descriptors.) Furthermore,
2883 errors from getrlimit() and setrlimit() were being ignored. Now they are
2884 logged to the main and panic logs, but Exim does carry on, to try to do its
2885 job under whatever limits there are.
2887 18. Imported PCRE 5.0.
2889 19. Trivial typo in log message " temporarily refused connection" (the leading
2892 20. If the log selector return_path_on_delivery was set and an address was
2893 redirected to /dev/null, the delivery process crashed because it assumed
2894 that a return path would always be set for a "successful" delivery. In this
2895 case, the whole delivery is bypassed as an optimization, and therefore no
2898 21. Internal re-arrangement: the function for sending a challenge and reading
2899 a response while authentication was assuming a zero-terminated challenge
2900 string. It's now changed to take a pointer and a length, to allow for
2901 binary data in such strings.
2903 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2905 23. Exim was not respecting finduser_retries when seeking the login of the
2906 uid under which it was called; it was always trying 10 times. (The default
2907 setting of finduser_retries is zero.) Also, it was sleeping after the final
2908 failure, which is pointless.
2910 24. Implemented tls_on_connect_ports.
2912 25. Implemented acl_smtp_predata.
2914 26. If the domain in control=submission is set empty, Exim assumes that the
2915 authenticated id is a complete email address when it generates From: or
2916 Sender: header lines.
2918 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2919 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2920 chown and chgrp in /bin and hostname in /usr/bin.
2922 28. Exim was keeping the "process log" file open after each use, just as it
2923 does for the main log. This opens the possibility of it remaining open for
2924 long periods when the USR1 signal hits a daemon. Occasional processlog
2925 errors were reported, that could have been caused by this. Anyway, it seems
2926 much more sensible not to leave this file open at all, so that is what now
2929 29. The long-running daemon process does not normally write to the log once it
2930 has entered its main loop, and it closes the log before doing so. This is
2931 so that log files can straightforwardly be renamed and moved. However,
2932 there are a couple of unusual error situations where the daemon does write
2933 log entries, and I had neglected to close the log afterwards.
2935 30. The text of an SMTP error response that was received during a remote
2936 delivery was being truncated at 512 bytes. This is too short for some of
2937 the long messages that one sometimes sees. I've increased the limit to
2940 31. It is now possible to make retry rules that apply only when a message has a
2941 specific sender, in particular, an empty sender.
2943 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2944 it possible to be selective about when SMTP synchronization is enforced.
2946 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2948 32. Implemented hosts_connection_nolog.
2950 33. Added an ACL for QUIT.
2952 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2955 35. Added mailbox_size and mailbox_filecount to appendfile.
2957 36. Added control = no_multiline_responses to ACLs.
2959 37. There was a bug in the logic of the code that waits for the clock to tick
2960 in the case where the clock went backwards by a substantial amount such
2961 that the microsecond fraction of "now" was more than the microsecond
2962 fraction of "then" (but the whole seconds number was less).
2964 38. Added support for the libradius Radius client library this is found on
2965 FreeBSD (previously only the radiusclient library was supported).
2971 1. When certain lookups returned multiple values in the form name=value, the
2972 quoting of the values was not always being done properly. Specifically:
2973 (a) If the value started with a double quote, but contained no whitespace,
2975 (b) If the value contained whitespace other than a space character (i.e.
2976 tabs or newlines or carriage returns) it was not quoted.
2977 This fix has been applied to the mysql and pgsql lookups by writing a
2978 separate quoting function and calling it from the lookup code. The fix
2979 should probably also be applied to nisplus, ibase and oracle lookups, but
2980 since I cannot test any of those, I have not disturbed their existing code.
2982 2. A hit in the callout cache for a specific address caused a log line with no
2983 reason for rejecting RCPT. Now it says "Previous (cached) callout
2984 verification failure".
2986 3. There was an off-by-one bug in the queryprogram router. An over-long
2987 return line was truncated at 256 instead of 255 characters, thereby
2988 overflowing its buffer with the terminating zero. As well as fixing this, I
2989 have increased the buffer size to 1024 (and made a note to document this).
2991 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2992 when Exim is waiting for an SMTP response from a remote server, Exim
2993 restarts its select() call on the socket, thereby resetting its timeout.
2994 This is not a problem when such interrupts are rare. Somebody set up a cron
2995 job to run exiwhat every 2 minutes, which is less than the normal select()
2996 timeout (5 or 10 minutes). This meant that the select() timeout never
2997 kicked in because it was always reset. I have fixed this by comparing the
2998 time when an interrupt arrives with the time at the start of the first call
2999 to select(). If more time than the timeout has elapsed, the interrupt is
3000 treated as a timeout.
3002 5. Some internal re-factoring in preparation for the addition of Sieve
3003 extensions (by MH). In particular, the "personal" test is moved to a
3004 separate function, and given an option for scanning Cc: and Bcc: (which is
3005 not set for Exim filters).
3007 6. When Exim created an email address using the login of the caller as the
3008 local part (e.g. when creating a From: or Sender: header line), it was not
3009 quoting the local part when it contained special characters such as @.
3011 7. Installed new OpenBSD configuration files.
3013 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3014 try to make them clearer.
3016 9. Callout options, other than the timeout value, were being ignored when
3017 verifying sender addresses in header lines. For example, when using
3019 verify = header_sender/callout=no_cache
3021 the cache was (incorrectly) being used.
3023 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3024 only the Cygwin environment.
3026 11. When return_path_on_delivery was set as a log selector, if different remote
3027 addresses in the same message used different return paths and parallel
3028 remote delivery occurred, the wrong values would sometimes be logged.
3029 (Whenever a remote delivery process finished, the return path value from
3030 the most recently started remote delivery process was logged.)
3032 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3033 lines when AUTH and/or TLS are in use. This is the "received protocol"
3034 field. Exim used to use "asmtp" for authenticated SMTP, without any
3035 indication (in the protocol name) for TLS use. Now it follows the RFC and
3036 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3037 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3038 names appear in log lines as well as in Received: header lines.
3040 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3041 extensions, and comparison tests, and to fix some bugs.
3043 14. Changes to the "personal" filter test:
3045 (1) The test was buggy in that it was just doing the equivalent of
3046 "contains" tests on header lines. For example, if a user's address was
3047 anne@some.where, the "personal" test would incorrectly be true for
3049 To: susanne@some.where
3051 This test is now done by extracting each address from the header in turn,
3052 and checking the entire address. Other tests that are part of "personal"
3053 are now done using regular expressions (for example, to check local parts
3054 of addresses in From: header lines).
3056 (2) The list of non-personal local parts in From: addresses has been
3057 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3058 taken from the Sieve specification recommendations.
3060 (3) If the message contains any header line starting with "List-" it is
3061 treated as non-personal.
3063 (4) The test for "circular" in the Subject: header line has been removed
3064 because it now seems ill-conceived.
3066 15. Minor typos in src/EDITME comments corrected.
3068 16. Installed latest exipick from John Jetmore.
3070 17. If headers_add on a router specified a text string that was too long for
3071 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3072 of string_sprintf() is now avoided.
3074 18. $message_body_size was not set (it was always zero) when running the DATA
3075 ACL and the local_scan() function.
3077 19. For the "mail" command in an Exim filter, no default was being set for
3078 the once_repeat time, causing a random time value to be used if "once" was
3079 specified. (If the value happened to be <= 0, no repeat happened.) The
3080 default is now 0s, meaning "never repeat". The "vacation" command was OK
3081 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3082 (I found it when inspecting the code).
3084 20. There is now an overall timeout for performing a callout verification. It
3085 defaults to 4 times the callout timeout, which applies to individual SMTP
3086 commands during the callout. The overall timeout applies when there is more
3087 than one host that can be tried. The timeout is checked before trying the
3088 next host. This prevents very long delays if there are a large number of
3089 hosts and all are timing out (e.g. when the network connections are timing
3090 out). The value of the overall timeout can be changed by specifying an
3091 additional sub-option for "callout", called "maxwait". For example:
3093 verify = sender/callout=5s,maxwait=20s
3095 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3096 to the end before writing, but this should make it even safer).
3098 22. Exim was forgetting that it had advertised PIPELINING for the second and
3099 subsequent messages on an SMTP connection. It was also not resetting its
3100 memory on STARTTLS and an internal HELO.
3102 23. When Exim logs an SMTP synchronization error within a session, it now
3103 records whether PIPELINING has been advertised or not.
3105 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3106 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3107 rather than long int.
3109 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3111 26. Added the never_mail option to autoreply.
3117 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3118 crash if the getsockname() call failed; this can happen if a connection is
3119 closed very soon after it is established. The problem was simply in the
3120 order in which certain operations were done, causing Exim to try to write
3121 to the SMTP stream before it had set up the file descriptor. The bug has
3122 been fixed by making things happen in the correct order.
3128 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3129 before the connection was closed, thus losing the rejection response.
3131 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3132 some early Solaris releases, but causes trouble in current releases where
3133 socklen_t is defined.
3135 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3138 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3141 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3142 that releasing the top of it at the end releases what was used for sub-
3143 expansions (unless the block got too big). However, discard this block if
3144 the first thing is a variable or header, so that we can use its block when
3145 it is dynamic (useful for very large $message_headers, for example).
3147 6. Lookups now cache *every* query, not just the most recent. A new, separate
3148 store pool is used for this. It can be recovered when all lookup caches are
3149 flushed. Lookups now release memory at the end of their result strings.
3150 This has involved some general refactoring of the lookup sources.
3152 7. Some code has been added to the store_xxx() functions to reduce the amount
3153 of flapping under certain conditions.
3155 8. log_incoming_interface used to affect only the <= reception log lines. Now
3156 it causes the local interface and port to be added to several more SMTP log
3157 lines, for example "SMTP connection from", and rejection lines.
3159 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3161 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3163 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3164 could be overwritten at the end of the current message (or the start of a
3165 new message if it was set in a HELO ACL). The value is now preserved for
3166 the duration of the SMTP connection.
3168 12. If a transport had a headers_rewrite setting, and a matching header line
3169 contained an unqualified address, that address was qualified, even if it
3170 did not match any rewriting rules. The underlying bug was that the values
3171 of the flags that permit the existence of unqualified sender and recipient
3172 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3173 non-local messages, and by -bnq for local messages) were not being
3174 preserved with the message after it was received.
3176 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3177 "next input=" as part of the text comprising the host identity instead of
3178 the correct text. The code was using the same buffer for two different
3179 strings. However, depending on which order the printing function evaluated
3180 its arguments, the bug did not always show up. Under Linux, for example, my
3181 test suite worked just fine.
3183 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3184 doesn't work with some older versions of Perl. It has been changed to "my",
3185 which in any case is probably the better facility to use.
3187 15. A really picky compiler found some instances of statements for creating
3188 error messages that either had too many or two few arguments for the format
3191 16. The size of the buffer for calls to the DNS resolver has been increased
3192 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3193 for addresses that have a lot of PTR records. This alleviates a problem; it
3194 does not fully solve it.
3196 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3197 buffer now truncates the list and logs the incident, which is the same
3198 action as happens when Exim is looking up a host name and its aliases.
3199 Previously in this situation something unpredictable would happen;
3200 sometimes it was "internal error: store_reset failed".
3202 18. If a server dropped the connection unexpectedly when an Exim client was
3203 using GnuTLS and trying to read a response, the client delivery process
3204 crashed while trying to generate an error log message.
3206 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3207 string, for example:
3209 warn message = H1: something\nH2: something
3211 the text was added as a single header line from Exim's point of view
3212 though it ended up OK in the delivered message. However, searching for the
3213 second and subsequent header lines using $h_h2: did not work. This has been
3214 fixed. Similarly, if a system filter added multiple headers in this way,
3215 the routers could not see them.
3217 20. Expanded the error message when iplsearch is called with an invalid key to
3218 suggest using net-iplsearch in a host list.
3220 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3221 ACLs are no longer actually imposed (and a message to that effect is
3224 22. If a "gecos" field in a passwd entry contained escaped characters, in
3225 particular, if it contained a \" sequence, Exim got it wrong when building
3226 a From: or a Sender: header from that name. A second bug also caused
3227 incorrect handling when an unquoted " was present following a character
3228 that needed quoting.
3230 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3231 was not being matched caselessly.
3233 24. Arranged for all hyphens in the exim.8 source to be escaped with
3236 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3237 properly. Recipient callout cache records were still being keyed to include
3238 the sender, even when use_sender was set false. This led to far more
3239 callouts that were necessary. The sender is no longer included in the key
3240 when use_sender is false.
3242 26. Added "control = submission" modifier to ACLs.
3244 27. Added the ${base62d: operator to decode base 62 numbers.
3246 28. dnsdb lookups can now access SRV records.
3248 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3249 the configuration file.
3251 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3252 (-v) mode. This makes the output for a verbose queue run more intelligible.
3254 31. Added a use_postmaster feature to recipient callouts.
3256 32. Added the $body_zerocount variable, containing the number of binary zero
3257 bytes in the message body.
3259 33. The time of last modification of the "new" subdirectory is now used as the
3260 "mailbox time last read" when there is a quota error for a maildir
3263 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3265 35. Added +ignore_unknown as a special item in host lists.
3267 36. Code for decoding IPv6 addresses in host lists is now included, even if
3268 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3269 address was recognized as an IP address, but was then not correctly decoded
3270 into binary, causing unexpected and incorrect effects when compared with
3277 1. Very minor rewording of debugging text in manualroute to say "list of
3278 hosts" instead of "hostlist".
3280 2. If verify=header_syntax was set, and a header line with an unqualified
3281 address (no domain) and a large number of spaces between the end of the
3282 name and the colon was received, the reception process suffered a buffer
3283 overflow, and (when I tested it) crashed. This was caused by some obsolete
3284 code that should have been removed. The fix is to remove it!
3286 3. When running in the test harness, delay a bit after writing a bounce
3287 message to get a bit more predictability in the log output.
3289 4. Added a call to search_tidyup() just before forking a reception process. In
3290 theory, someone could use a lookup in the expansion of smtp_accept_max_
3291 per_host which, without the tidyup, could leave open a database connection.
3293 5. Added the variables $recipient_data and $sender_data which get set from a
3294 lookup success in an ACL "recipients" or "senders" condition, or a router
3295 "senders" option, similar to $domain_data and $local_part_data.
3297 6. Moved the writing of debug_print from before to after the "senders" test
3300 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3301 problems for message scanning, either using a data ACL, or using
3302 local_scan() because the Received: header was not generated till after they
3303 were called (in order to set the time as the time of reception completion).
3304 I have revised the way this works. The header is now generated after the
3305 body is received, but before the ACL or local_scan() are called. After they
3306 are run, the timestamp in the header is updated.
3312 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3313 before starting a queue runner without re-exec. This happened only when
3314 deliver_drop_privilege was set or when the Exim user was set to root. The
3315 effect of the bug was that timeouts during subsequent deliveries caused
3316 crashes instead of being properly handled. The handler is now left at its
3317 default (and expected) setting.
3319 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3320 message, again when deliver_drop_privilege is set or Exim is run as root.
3321 The bug described in (1) was not present in this case, but the tidying up
3322 of the other signals was missing. I have made the two cases consistent.
3324 3. The ignore_target_hosts setting on a manualroute router was being ignored
3325 for hosts that were looked up using the /MX notation.
3327 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3330 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3331 operated on the sender address. After changing the $sender_address to <>
3332 for the sender address verify, Exim was re-instated it as the original
3333 (before rewriting) address, but remembering that it had rewritten it, so it
3334 wasn't rewriting it again. This bug also had the effect of breaking the
3335 sender address verification caching when the sender address was rewritten.
3337 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3338 This has been changed so that if the ip literal address matches
3339 ignore_target_hosts, the router declines.
3341 7. Added expansion conditions match_domain, match_address, and match_local_
3342 part (NOT match_host).
3344 8. The placeholder for the Received: header didn't have a length field set.
3346 9. Added code to Exim itself and to exim_lock to test for a specific race
3347 condition that could lead to file corruption when using MBX delivery. The
3348 issue is with the lockfile that is created in /tmp. If this file is removed
3349 after a process has opened it but before that process has acquired a lock,
3350 there is the potential for a second process to recreate the file and also
3351 acquire a lock. This could lead to two Exim processes writing to the file
3352 at the same time. The added code performs the same test as UW imapd; it
3353 checks after acquiring the lock that its file descriptor still refers to
3354 the same named file.
3356 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3357 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3360 11. Added the smtp_active_hostname option. If used, this will typically be made
3361 to depend on the incoming interface address. Because $interface_address is
3362 not set up until the daemon has forked a reception process, error responses
3363 that can happen earlier (such as "too many connections") no longer contain
3366 12. If an expansion in a condition on a "warn" statement fails because a lookup
3367 defers, the "warn" statement is abandoned, and the next ACL statement is
3368 processed. Previously this caused the whole ACL to be aborted.
3370 13. Added the iplsearch lookup type.
3372 14. Added ident_timeout as a log selector.
3374 15. Added tls_certificate_verified as a log selector.
3376 16. Added a global option tls_require_ciphers (compare the smtp transport
3377 option of the same name). This controls incoming TLS connections.
3379 17. I finally figured out how to make tls_require_ciphers do a similar thing
3380 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3381 before starting the TLS session.
3383 18. Tabs are now shown as \t in -bP output.
3385 19. If the log selector return_path_on_delivery was set, Exim crashed when
3386 bouncing a message because it had too many Received: header lines.
3388 20. If two routers both had headers_remove settings, and the first one included
3389 a superfluous trailing colon, the final name in the first list and the
3390 first name in the second list were incorrectly joined into one item (with a
3391 colon in the middle).
3397 1. Added -C and -D options to the exinext utility, mainly to make it easier
3398 to include in the automated testing, but these could be helpful when
3399 multiple configurations are in use.
3401 2. The exinext utility was not formatting the output nicely when there was
3402 an alternate port involved in the retry record key, nor when there was a
3403 message id as well (for retries that were specific to a specific message
3404 and a specific host). It was also confused by IPv6 addresses, because of
3405 the additional colons they contain. I have fixed the IPv4 problem, and
3406 patched it up to do a reasonable job for IPv6.
3408 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3409 delivery, the log line now contains "pipelined" if PIPELINING was used.
3411 4. An SMTP transport process used to panic and die if the bind() call to set
3412 an explicit outgoing interface failed. This has been changed; it is now
3413 treated in the same way as a connect() failure.
3415 5. A reference to $sender_host_name in the part of a conditional expansion
3416 that was being skipped was still causing a DNS lookup. This no longer
3419 6. The def: expansion condition was not recognizing references to header lines
3420 that used bh_ and bheader_.
3422 7. Added the _cache feature to named lists.
3424 8. The code for checking quota_filecount in the appendfile transport was
3425 allowing one more file than it should have been.
3427 9. For compatibility with Sendmail, the command line option
3435 and sets the incoming protocol and host name (for trusted callers). The
3436 host name and its colon can be omitted when only the protocol is to be set.
3437 Note the Exim already has two private options, -pd and -ps, that refer to
3438 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3439 "s", but I don't think that's a major issue.
3441 10. A number of refactoring changes to the code, none of which should affect
3444 (a) The number of logging options was getting close to filling up the
3445 32-bit word that was used as a bit map. I have split them into two classes:
3446 those that are passed in the argument to log_write(), and those that are
3447 only ever tested independently outside of that function. These are now in
3448 separate 32-bit words, so there is plenty of room for expansion again.
3449 There is no change in the user interface or the logging behaviour.
3451 (b) When building, for example, log lines, the code previously used a
3452 macro that called string_cat() twice, in order to add two strings. This is
3453 not really sufficiently general. Furthermore, there was one instance where
3454 it was actually wrong because one of the argument was used twice, and in
3455 one call a function was used. (As it happened, calling the function twice
3456 did not affect the overall behaviour.) The macro has been replaced by a
3457 function that can join an arbitrary number of extra strings onto a growing
3460 (c) The code for expansion conditions now uses a table and a binary chop
3461 instead of a serial search (which was left over from when there were very
3462 few conditions). Also, it now recognizes conditions like "pam" even when
3463 the relevant support is not compiled in: a suitably worded error message is
3464 given if an attempt is made to use such a condition.
3466 11. Added ${time_interval:xxxxx}.
3468 12. A bug was causing one of the ddress fields not to be passed back correctly
3469 from remote delivery subprocesses. The field in question was not being
3470 subsequently used, so this caused to problems in practice.
3472 13. Added new log selectors queue_time and deliver_time.
3474 14. Might have fixed a bug in maildirsizefile handling that threw up
3475 "unexpected character" debug warnings, and recalculated the data
3476 unnecessarily. In any case, I expanded the warning message to give more
3479 15. Added the message "Restricted characters in address" to the statements in
3480 the default ACL that block characters like @ and % in local parts.
3482 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3483 Three changes have been made:
3485 (a) There was a serious bug; a negative response to MAIL caused the whole
3486 recipient domain to be cached as invalid, thereby blocking all messages
3487 to all local parts at the same domain, from all senders. This bug has
3488 been fixed. The domain is no longer cached after a negative response to
3489 MAIL if the sender used is not empty.
3491 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3494 (c) A new callout option, "use_sender" has been added for people who want
3495 the modified behaviour.
3501 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3504 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3505 indeed breaks things for older releases.
3507 3. Added additional logging to the case where there is a problem reading data
3508 from a filter that is running in a subprocess using a pipe, in order to
3509 try to track down a specific problem.
3511 4. Testing facility fudge: when running in the test harness and attempting
3512 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3513 getting "No route to host". Convert this to a timeout.
3515 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3518 6. Some OS don't have socklen_t but use size_t instead. This affects the
3519 fifth argument of getsockopt() amongst other things. This is now
3520 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3521 can be set for individual OS. I have set it for SunOS5, OSF1, and
3522 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3523 some earlier ones do not.
3525 7. Change 4.30/15 was not doing the test caselessly.
3527 8. The standard form for an IPv6 address literal was being rejected by address
3528 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3529 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3530 this, as well as the form without the "IPv6" on the front (but only when
3531 address literals are enabled, of course).
3533 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3535 10. Exim crashed if a message with an empty sender address specified by -f
3536 encountered a router with an errors_to setting. This could be provoked only
3537 by a command such as
3541 where an empty string was supplied; "<>" did not hit this bug.
3543 11. Installed PCRE release 4.5.
3545 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3546 remained set. It is now erased.
3548 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3549 times from message ids (which are base 36 rather than the normal 62).
3551 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3552 were being counted as actual protocol errors, and logged if the log
3553 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3554 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3555 and DATA following a set of rejected RCPTs do not count as protocol errors.
3556 In other words, Exim assumes they were pipelined, though this may not
3557 actually be the case. Of course, in all cases the client gets an
3558 appropriate error code.
3560 15. If a lookup fails in an ACL condition, a message about the failure may
3561 be available; it is used if testing the ACL cannot continue, because most
3562 such messages specify what the cause of the deferral is. However, some
3563 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3564 that caused an old message to be retained and used if a later statement
3565 caused a defer, replacing the real cause of the deferral.
3567 16. If an IP address had so many PTR records that the DNS lookup buffer
3568 was not large enough to hold them, Exim could crash while trying to process
3569 the truncated data. It now detects and logs this case.
3571 17. Further to 4.21/58, another change has been made: if (and only if) the
3572 first line of a message (the first header line) ends with CRLF, a bare LF
3573 in a subsequent header line has a space inserted after it, so as not to
3574 terminate the header.
3576 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3577 unnecessarily, used atoi() instead of strtol(), and didn't check the
3578 termination when getting file sizes from file names by regex.
3580 19. Completely re-implemented the support for maildirsize files, in the light
3581 of a number of problems with the previous contributed implementation
3582 (4.30/29). In particular:
3584 . If the quota is zero, the maildirsize file is maintained, but no quota is
3587 . If the maildir directory does not exist, it is created before any attempt
3588 to write a maildirsize file.
3590 . The quota value in the file is just a cache; if the quota is changed in
3591 the transport, the new value overrides.
3593 . A regular expression is available for excluding directories from the
3596 20. The autoreply transport checks the characters in options that define the
3597 message's headers; it allows continued headers, but it was checking with
3598 isspace() after an embedded newline instead of explicitly looking for a
3601 21. If all the "regular" hosts to which an address was routed had passed their
3602 expiry times, and had not reached their retry times, the address was
3603 bounced, even if fallback hosts were defined. Now Exim should go on to try
3606 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3607 equivalent code in the SMTP transport. Some hosts send humungous responses
3608 to HELO/EHLO, more than 1024 it seems.
3610 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3611 gives compiler warnings in some environments. I've now done it "properly",
3614 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3615 (because of problems with the built-in one) was declared to return uschar *
3616 instead of char *, causing compiler failure.
3618 25. Fixed a file descriptor leak when processing alias/forward files.
3620 26. Fixed a minor format string issue in dbfn.c.
3622 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3624 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3625 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3628 29. When a sender address is verified, it is cached, to save repeating the test
3629 when there is more than one recipient in a message. However, when the
3630 verification involves a callout, it is possible for different callout
3631 options to be set for different recipients. It is too complicated to keep
3632 track of this in the cache, so now Exim always runs a verification when a
3633 callout is required, relying on the callout cache for the optimization.
3634 The overhead is duplication of the address routing, but this should not be
3637 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3638 to be verified with callout=postmaster, and the main callout worked but the
3639 postmaster check failed, the verification correctly failed. However, if a
3640 subsequent RCPT command asked for sender verification *without* the
3641 postmaster check, incorrect caching caused this verification also to fail,
3644 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3645 it was not caching the DNS options (qualify_single, search_parents) that
3646 were used when the lookup failed. A subsequent lookup with different
3647 options therefore always gave the same answer, though there were cases
3648 where it should not have. (Example: a "domains = !$mx_any" option on a
3649 dnslookup router: the "domains" option is always processed without any
3650 widening, but the router might have qualify_single set.) Now Exim uses the
3651 cached value only when the same options are set.
3653 32. Added John Jetmore's "exipick" utility to the distribution.
3655 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3656 than a timeout (e.g. a certificate is required, and is not provided), an
3657 Exim server now closes the connection immediately. Previously it waited for
3658 the client to close - but if the client is SSL, it seems that they each
3659 wait for each other, leading to a delay before one of them times out.
3661 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3662 maintained 0.8.x compatibility because I don't think many are using it, and
3663 it is clearly obsolete.
3665 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3668 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3669 client certificate was expired. A simple patch fixes this, though I don't
3670 understand the full logic of why the verify callback is called multiple
3673 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3674 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3675 which causes problems with some clients (such as the Certicom SSL Plus
3676 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3677 disables the coutermeasure allowing Eudora to connect."
3679 38. Exim was not checking that a write() to a log file succeeded. This could
3680 lead to Bad Things if a log got too big, in particular if it hit a file
3681 size limit. Exim now panics and dies if it cannot write to a log file, just
3682 as it does if it cannot open a log file.
3684 39. Modified OS/Makefile-Linux so that it now contains
3686 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3688 The two -D definitions ensure that Exim is compiled with large file
3689 support, which makes it possible to handle log files that are bigger than
3692 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3693 instance) a domain was checked against a named list that involved a lookup,
3694 causing $domain_data to be set, then another domain was checked against the
3695 same list, then the first domain was re-checked, the value of $domain_data
3696 after the final check could be wrong. In particular, if the second check
3697 failed, it could be set empty. This bug probably also applied to
3700 41. The strip_trailing_dot option was not being applied to the address given
3701 with the -f command-line option.
3703 42. The code for reading a message's header from the spool was incrementing
3704 $received_count, but never initializing it. This meant that the value was
3705 incorrect (doubled) while delivering a message in the same process in which
3706 it was received. In the most common configuration of Exim, this never
3707 happens - a fresh exec is done - but it can happen when
3708 deliver_drop_privilege is set.
3710 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3711 it now includes up to 150 characters of the unexpected data in the log
3714 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3715 and building data strings. The size of both of these buffers was 10 000
3716 bytes - far larger than anybody would *ever* want, thought I. Needless to
3717 say, somebody hit the limit. I have increased the maximum line length to
3718 20 000 and the maximum data length of concatenated lines to 100 000. I have
3719 also fixed two bugs, because there was no checking on these buffers. Tsk,
3720 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3721 buffer is too small.
3723 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3724 lsearch lookups. Now it does.
3726 46. When parsing a route_list item in a manualroute router, a fixed-length
3727 buffer was used for the list of hosts. I made this 1024 bytes long,
3728 thinking that nobody would ever have a list of hosts that long. Wrong.
3729 Somebody had a whole pile of complicated expansion conditions, and the
3730 string was silently truncated, leading to an expansion error. It turns out
3731 that it is easier to change to an unlimited length (owing to other changes
3732 that have happened since this code was originally written) than to build
3733 structure for giving a limitation error. The length of the item that
3734 expands into the list of hosts is now unlimited.
3736 47. The lsearch lookup could not handle data where the length of text line was
3737 more than 4095 characters. Such lines were truncated, leading to shortened
3738 data being returned. It should now handle lines of any length.
3740 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3741 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3744 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3745 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3746 no longer happens in comment lines. A list of replacements is now placed
3747 at the head of all of the source files, except those whose only change is
3748 to replace PERL_COMMAND in the very first #! line.
3750 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3751 messages on the queue, with a bottom-up merge sort, using code contributed
3752 by Michael Haardt. This should make operations like -bp somewhat faster on
3753 large queues. It won't affect queue runners, except when queue_run_in_order
3756 51. Installed eximstats 1.31 in the distribution.
3758 52. Added support for SRV lookups to the dnslookup router.
3760 53. If an ACL referred to $message_body or $message_body_end, the value was not
3761 reset for any messages that followed in the same SMTP session.
3763 54. The store-handling optimization for building very long strings was not
3764 differentiating between the different store pools. I don't think this
3765 actually made any difference in practice, but I've tidied it.
3767 55. While running the routers to verify a sender address, $sender_address
3768 was still set to the sender address. This is wrong, because when routing to
3769 send a bounce to the sender, it would be empty. Therefore, I have changed
3770 it so that, while verifying a sender address, $sender_address is set to <>.
3771 (There is no change to what happens when verifying a recipient address.)
3773 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3774 target A or AAAA records (if not already returned) without resetting the
3775 qualify_single or search_parents options of the DNS resolver. These are
3776 inappropriate in this case because the targets of MX and SRV records must
3777 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3778 target that, when qualified, matched something in the local domain. These
3779 two options are now turned off when doing these lookups.
3781 57. It seems that at least some releases of Reiserfs (which does not have the
3782 concept of a fixed number of inodes) returns zero and not -1 for the
3783 number of available inodes. This interacted badly with check_spool_inodes,
3784 which assumed that -1 was the "no such thing" setting. What I have done is
3785 to check that the total number of inodes is greater than zero before doing
3786 the test of how many are available.
3788 58. When a "warn" ACL statement has a log_message modifier, the message is
3789 remembered, and not repeated. This is to avoid a lot of repetition when a
3790 message has many recipients that cause the same warning to be written.
3791 Howewer, Exim was preserving the list of already written lines for an
3792 entire SMTP session, which doesn't seem right. The memory is now reset if a
3793 new message is started.
3795 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3796 debugging output unless log_rewrite was also set.
3798 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3799 of (int)(handle) when we know that handle contains (void *)(-1).
3801 61. The Exim daemon panic-logs an error return when it closes the incoming
3802 connection. However "connection reset by peer" seems to be common, and
3803 isn't really an error worthy of noting specially, so that particular error
3806 62. When Exim is trying to find all the local interfaces, it used to panic and
3807 die if the ioctl to get the interface flags failed. However, it seems that
3808 on at least one OS (Solaris 9) it is possible to have an interface that is
3809 included in the list of interfaces, but for which you get a failure error
3810 for this call. This happens when the interface is not "plumbed" into a
3811 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3812 failure of the "get flags" call assumes that the interface is down.
3814 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3815 makes life easier for people who are doing arithmetic on fields extracted
3816 from dates, where you often get leading zeros that should not be
3817 interpreted as octal.
3819 64. Added qualify_domain to the redirect router, to override the global
3822 65. If a pathologically long header line contained very many addresses (the
3823 report of this problem mentioned 10 000) and each of them was rewritten,
3824 Exim could use up a very large amount of memory. (It kept on making new
3825 copies of the header line as it rewrote, and never released the old ones.)
3826 At the expense of a bit more processing, the header rewriting function has
3827 been changed so that it no longer eats memory in this way.
3829 66. The generation of the Received: header has been moved from the time that a
3830 message starts to be received, to the time that it finishes. The timestamp
3831 in the Received: header should now be very close to that of the <= log
3832 line. There are two side-effects of this change:
3834 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3835 logged header lines no longer include the local Received: line, because
3836 it has not yet been created. The same applies to a copy of the message
3837 that is returned to a non-SMTP sender when a message is rejected.
3839 (b) When a filter file is tested using -bf, no additional Received: header
3840 is added to the test message. After some thought, I decided that this
3843 This change does not affect the value of $received_for. It is still set
3844 after address rewriting, but before local_scan() is called.
3846 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3848 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3849 gave an unhelpful panic error message, and a defer error. I have managed to
3850 change this behaviour so that it now rejects any supplied certificate,
3851 which seems right, as the list of acceptable certificates is empty.
3853 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3854 gave an unhelpful defer error. I have not managed to make this reject any
3855 supplied certificates, but the error message it gives is "no certificate
3856 supplied", which is not helpful.
3858 70. exigrep's output now also includes lines that are not associated with any
3859 message, but which match the given pattern. Implemented by a patch from
3860 Martin Sluka, which also tidied up the Perl a bit.
3862 71. Recipient callout verification, like sender verification, was using <> in
3863 the MAIL FROM command. This isn't really the right thing, since the actual
3864 sender may affect whether the remote host accepts the recipient or not. I
3865 have changed it to use the actual sender in the callout; this means that
3866 the cache record is now keyed on a recipient/sender pair, not just the
3867 recipient address. There doesn't seem to be a real danger of callout loops,
3868 since a callout by the remote host to check the sender would use <>.
3869 [SEE ABOVE: changed after hitting problems.]
3871 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3872 temporary errors. However, in the case of such a code being given after
3873 the end of a data transmission (i.e. after ".") Exim was failing to write
3874 a retry record for the message. (Yes, there was some broken host that was
3875 actually sending 8xx at this point.)
3877 73. An unknown lookup type in a host list could cause Exim to panic-die when
3878 the list was checked. (An example that provoked this was putting <; in the
3879 middle of a list instead of at the start.) If this happened during a DATA
3880 ACL check, a -D file could be left lying around. This kind of configuration
3881 error no longer causes Exim to die; instead it causes a defer errror. The
3882 incident is still logged to the main and panic logs.
3884 74. Buglet left over from Exim 3 conversion. The message "too many messages
3885 in one connection" was written to the rejectlog but not the mainlog, except
3886 when address rewriting (yes!) was being logged.
3888 75. Added write_rejectlog option.
3890 76. When a system filter was run not as root (that is, when system_filter_user
3891 was set), the values of the $n variables were not being returned to the
3892 main process; thus, they were not subsequently available in the $sn
3895 77. Added +return_path_on_delivery log selector.
3897 78. A connection timeout was being treated differently from recipients deferred
3898 when testing hosts_max_try with a message that was older than the host's
3899 retry timeout. (The host should not be counted, thus allowing all hosts to
3900 be tried at least once before bouncing.) This may have been the cause of an
3901 occasionally reported bug whereby a message would remain on the queue
3902 longer than the retry timeout, but would be bounced if a delivery was
3903 forced. I say "may" because I never totally pinned down the problem;
3904 setting up timeout/retry tests is difficult. See also the next item.
3906 79. The ultimate address timeout was not being applied to errors that involved
3907 a combination of host plus message (for example, a timeout on a MAIL
3908 command). When an address resolved to a number of possible hosts, and they
3909 were not all tried for each delivery (e.g. because of hosts_max_try), a
3910 message could remain on the queue longer than the retry timeout.
3912 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3915 81. Fixed an obscure SMTP outgoing bug which required at least the following
3916 conditions: (a) there was another message waiting for the same server;
3917 (b) the server returned 5xx to all RCPT commands in the first message so
3918 that the message was not completed; (c) the server dropped the connection
3919 or gave a negative response to the RSET that Exim sends to abort the
3920 transaction. The observed case was a dropped connection after DATA that had
3921 been sent in pipelining mode. That is, the server had advertised PIPELINING
3922 but was not implementing it correctly. The effect of the bug was incorrect
3923 behaviour, such as trying another host, and this could lead to a crash.
3929 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3930 and daemon.c were passed as pointers to ints; they should have been
3931 pointers to socklen_t variables (which are typically unsigned ints).
3933 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3936 3. Fixed a really odd bug that affected only the testing scheme; patching a
3937 certain fixed string in the binary changed the value of another string that
3938 happened to be identical to the end of the original first string.
3940 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3941 name", it returns that address as the IP address. On some operating
3942 systems (e.g. Solaris), it also passes back the IP address string as the
3943 "host name". However, on others (e.g. Linux), it passes back an empty
3944 string. Exim wasn't checking for this, and was changing the host name to an
3945 empty string, assuming it had been canonicized.
3947 5. Although rare, it is permitted to have more than one PTR record for a given
3948 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3949 all the names associated with an address, because they do in Solaris.
3950 However, it seems that they do not in Linux for data that comes from the
3951 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3952 I found this out when I moved to a new Linux workstation and tried to run
3953 the Exim test suite.
3955 To get round this problem I have changed the code so that it now does its
3956 own call to the DNS to look up PTR records when searching for a host name.
3957 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3958 addresses that are only in /etc/hosts are still found.
3960 This behaviour is, however, controlled by an option called host_lookup_
3961 order, which defaults to "bydns:byaddr". If people want to use the other
3962 order, or indeed, just use one or the other means of lookup, they can
3963 specify it in this variable.
3965 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3966 some operating systems, this comes back from gethostbyaddr() as an empty
3967 string, and this is what Exim used to test for. However, it seems that in
3968 other systems, "." is yielded. Exim now tests for this case too.
3970 7. The values of check_spool_space and check_log_space are now held internally
3971 as a number of kilobytes instead of an absolute number of bytes. If a
3972 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3973 kilobyte. This means that much larger values can be stored.
3975 8. Exim monitor: an attempt to get the action menu when not actually pointing
3976 at a message produces an empty menu entitled "No message selected". This
3977 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3978 no entries in it ("Shell widget menu has zero width and/or height"). So I
3979 have added a single, blank menu entry in this case.
3981 9. Added ${quote_local_part.
3983 10. MIME decoding is now applied to the contents of Subject: header lines when
3986 11. Now that a reference to $sender_host_address automatically causes a reverse
3987 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3988 host lookup before query-style lookups in lists that might use this
3989 variable. This has therefore been abolished, and the "net-" prefix is no
3990 longer necessary for query-style lookups.
3992 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3993 have been a typo for LFLAGS, so it has been changed.
3995 13. The install script calls Exim with "-C /dev/null" in order to find the
3996 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3997 to be output. Howeve, since Exim outputs its version number before the
3998 error, it didn't break the script. It just looked ugly. I fixed this by
3999 always allowing "-C /dev/null" if the caller is root.
4001 14. Ignore overlarge ACL variable number when reading spool file - insurance
4002 against a later release with more variables having written the file.
4004 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4005 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4006 this, as well as the form without the "IPv6" on the front.
4008 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4009 OS/Makefile-Darwin file.
4011 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4012 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4014 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4015 variable. However, some OS use ints for inodes. I've added cast to long int
4016 to get rid of the compiler warning.
4018 19. I had forgotten to lock out "/../" in configuration file names when
4019 ALT_CONFIG_PREFIX was set.
4021 20. Routers used for verification do not need to specify transports. However,
4022 if such a router generated a host list, and callout was configured, Exim
4023 crashed, because it could not find a port number from the (non-existent)
4024 transport. It now assumes port 25 in this circumstance.
4026 21. Added the -t option to exigrep.
4028 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4029 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4030 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4031 them is set, LOOKUP_LSEARCH is forced.
4033 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4035 24. Added sender and host information to the "rejected by local_scan()" log
4036 line; previously there was no indication of these.
4038 25. Added .include_if_exists.
4040 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4041 Linux. It turns out that not all file systems support this. Apparently some
4042 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4043 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4044 not supported on the file descriptor, is now ignored when Exim is trying to
4045 sync a directory. This applies only to Linux.
4047 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4049 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4050 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4053 29. Added support for maildirsize files from supplied patch (modified a bit).
4055 30. The use of :fail: followed by an empty string could lead Exim to respond to
4056 sender verification failures with (e.g.):
4058 550 Verification failed for <xxx>
4059 550 Sender verify failed
4061 where the first response line was missing the '-' that indicates it is not
4062 the final line of the response.
4064 31. The loop for finding the name of the user that called Exim had a hardwired
4065 limit of 10; it now uses the value of finduser_retries, which is used for
4066 all other user lookups.
4068 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4071 33. Exim was neglecting to zero errno before one call of strtol() when
4072 expanding a string and expecting an integer value. On some systems this
4073 resulted in spurious "integer overflow" errors. Also, it was casting the
4074 result into an int without checking.
4076 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4077 did not work. The code looks as if it has *never* worked, though it appears
4078 to have been documented since at least releast 1.62. I have made it work.
4080 35. The "timeout_DNS" error in retry rules, also documented since at least
4081 1.62, also never worked. As it isn't clear exactly what this means, and
4082 clearly it isn't a major issue, I have abolished the feature by treating it
4083 as "timeout", and writing a warning to the main and panic logs.
4085 36. The display of retry rules for -brt wasn't always showing the error code
4088 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4089 timeout_connect_A, timeout_connect_MX.
4091 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4092 to the empty sender.
4094 39. The daemon was not analysing the content of -oX till after it had closed
4095 stderr and disconnected from the controlling terminal. This meant that any
4096 syntax errors were only noted on the panic log, and the return code from
4097 the command was 0. By re-arranging the code a little, I've made the
4098 decoding happen first, so such errors now appear on stderr, and the return
4099 code is 1. However, the actual setting up of the sockets still happens in
4100 the disconnected process, so errors there are still only recorded on the
4103 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4104 connections (as happens on some IP stacks) was logged at start up time as
4105 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4106 it from "IPv6 and IPv4", which means that two separate sockets are being
4109 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4110 says whether AF_INET or AF_INET6 was passed as an argument.
4112 42. Exiwhat output was messed up when time zones were included in log
4115 43. Exiwhat now gives more information about the daemon's listening ports,
4116 and whether -tls-on-connect was used.
4118 44. The "port" option of the smtp transport is now expanded.
4120 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4121 silently ignored. Now an error message is written to the main and panic
4124 46. There's a new ACL modifier called "logwrite" which writes to a log file
4125 as soon as it is encountered.
4127 47. Added $local_user_uid and $local_user_gid at routing time.
4129 48. Exim crashed when trying to verify a sender address that was being
4132 49. Exim was recognizing only a space character after ".include". It now also
4133 recognizes a tab character.
4135 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4136 extracting the relevant information from the specification. The man page no
4137 longer contains scrambled data for the -d option, and I've added a section
4138 at the front about calling Exim under different names.
4140 51. Added "extra_headers" argument to the "mail" command in filter files.
4142 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4145 53. Installed eximstats 1.29.
4147 54. Added transport_filter_timeout as a generic transport option.
4149 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4150 Cc: header lines. This was required by RFC 822, but it not required by RFC
4153 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4154 incoming messages that did not have them. Now it does so only if the
4155 message originates locally, that is, if there is no associated remote host
4156 address. When Resent- header lines are present, this applies to the Resent-
4157 lines rather than the non-Resent- lines.
4159 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4160 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4162 58. Messages for configuration errors now include the name of the main
4163 configuration file - useful now that there may be more than one file in a
4164 list (.included file names were always shown).
4166 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4167 for those rare installations that do not start the daemon as root or run it
4168 setuid root. I've cut out the call to initgroups() if the daemon is not
4171 60. The Exim user and group can now be bound into the binary as text strings
4172 that are looked up at the start of Exim's processing.
4174 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4176 62. Added $mailstore_basename variable.
4178 63. Installed patch to sieve.c from Michael Haardt.
4180 64. When Exim failed to open the panic log after failing to open the main log,
4181 the original message it was trying to log was written to stderr and debug
4182 output, but if they were not available (the usual case in production), it
4183 was lost. Now it is written to syslog before the two lines that record the
4184 failures to open the logs.
4186 65. Users' Exim filters run in subprocesses under the user's uid. It is
4187 possible for a "deliver" command or an alias in a "personal" command to
4188 provoke an address rewrite. If logging of address rewriting is configured,
4189 this fails because the process is not running as root or exim. There may be
4190 a better way of dealing with this, but for the moment (because 4.30 needs
4191 to be released), I have disabled address rewrite logging when running a
4192 filter in a non-root, non-exim process.
4198 1. The buildconfig auxiliary program wasn't quoting the value set for
4199 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4200 not defined. This bug was masked in 4.22 by the effect that was fixed in
4203 2. Some messages that were rejected after a message id was allocated were
4204 shown as "incomplete" by exigrep. It no longer does this for messages that
4205 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4207 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4208 have allow_domain_literals set, the ID did not get logged in the <= line.
4209 Domain literals are now always recognized in Message-ID: header lines.
4211 4. The first argument for a ${extract expansion item is the key name or field
4212 number. Leading and trailing spaces in this item were not being ignored,
4213 causing some misleading effects.
4215 5. When deliver_drop_privilege was set, single queue runner processes started
4216 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4217 same command in the process it spins off) were not dropping privilege.
4219 6. When the daemon running as "exim" started a queue runner, it always
4220 re-executed Exim in the spun-off process. This is a waste of effort when
4221 deliver_drop_privilege is set. The new process now just calls the
4222 queue-runner function directly.
4228 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4231 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4232 ignored. Though the use of -oP was forcing the writing of a pid file, it
4233 was always written to the default place.
4235 3. If the message "no IP address found for host xxxx" is generated during
4236 incoming verification, it is now followed by identification of the incoming
4237 connection (so you can more easily find what provoked it).
4239 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4241 5. Added some features to "harden" Exim a bit more against certain attacks:
4243 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4244 be put in Local/Makefile. This is like the never_users runtime option,
4245 but it cannot be overridden. The default setting is "root".
4247 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4248 prefix string with which any file named in a -C command line option
4251 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4252 is retained for -C and -D only if the caller of Exim is root. Without
4253 it, the exim user may also use -C and -D and retain privilege.
4255 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4256 command line option is disabled.
4258 6. Macro names set by the -D option must start with an upper case letter, just
4259 like macro names defined in the configuration file.
4261 7. Added "dereference=" facility to LDAP.
4263 8. Two instances of the typo "uknown" in the source files are fixed.
4265 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4266 the Configure-Makefile script screwed up while processing it.
4268 10. Incorporated PCRE 4.4.
4270 11. The SMTP synchronization check was not operating right at the start of an
4271 SMTP session. For example, it could not catch a HELO sent before the client
4272 waited for the greeting. There is now a check for outstanding input at the
4273 point when the greeting is written. Because of the duplex, asynchronous
4274 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4275 way, but not yet received, when the check is performed.
4277 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4278 on TCP/IP sockets, because this apparently causes some broken clients to
4281 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4282 unchanged) from the Cygwin maintainer.
4284 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4285 was supported instead of testing for mbx. Effectively a typo.
4287 15. The spa authenticator server code was not checking that the input it
4288 received was valid base64.
4290 16. The debug output line for the "set" modifier in ACLs was not showing the
4291 name of the variable that was being set.
4293 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4295 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4296 lookup on its own. Something else was needed to trigger the lookup. For
4297 example, a match in host_lookup or the need for a host name in a host list.
4298 Now, if $sender_host_name is referenced and the host name has not yet been
4299 looked up, a lookup is performed. If the lookup fails, the variable remains
4300 empty, and $host_lookup_failed is set to "1".
4302 19. Added "eqi" as a case-independent comparison operator.
4304 20. The saslauthd authentication condition could segfault if neither service
4305 nor realm was specified.
4307 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4308 error message that was logged was misleading, and incoming SMTP
4309 connections were dropped. The message is now more accurate, and temporary
4310 errors are given to SMTP connections.
4312 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4313 (except RSET) with a 421 error, until QUIT is received. However, it was
4314 failing to send a response to QUIT.
4316 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4317 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4318 ACL is now run after the helo_try_verify_hosts code.
4320 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4321 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4322 case-independent, so other case variants are also recognized). Apparently
4323 some systems use these upper case variants.
4325 25. If more than two messages were waiting for the same host, and a transport
4326 filter was specified for the transport, Exim sent two messages over the
4327 same TCP/IP connection, and then failed with "socket operation on non-
4328 socket" when it tried to send the third.
4330 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4332 27. The extern definition of crypt16() in expand.c was not being excluded when
4333 the OS had its own crypt16() function.
4335 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4336 as a preferred synonym for return_size_limit, both as an option and as an
4339 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4341 30. Changed the default configuration ACL to relax the local part checking rule
4342 for addresses that are not in any local domains. For these addresses,
4343 slashes and pipe symbols are allowed within local parts, but the sequence
4344 /../ is explicitly forbidden.
4346 31. SPA server authentication was not clearing the challenge buffer before
4349 32. log_message in a "warn" ACL statement was writing to the reject log as
4350 well as to the main log, which contradicts the documentation and doesn't
4351 seem right (because no rejection is happening). So I have stopped it.
4353 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4354 However, I am unable to do any testing of this.
4356 34. Fixed an infelicity in the appendfile transport. When checking directories
4357 for a mailbox, to see if any needed to be created, it was accidentally
4358 using path names with one or more superfluous leading slashes; tracing
4359 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4361 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4362 contents are added to the log line that is written for every discarded
4363 recipient. (Previously a log_message setting was ignored.)
4365 36. The ${quote: operator now quotes the string if it is empty.
4367 37. The install script runs exim in order to find its version number. If for
4368 some reason other than non-existence or emptiness, which it checks, it
4369 could not run './exim', it was installing it with an empty version number,
4370 i.e. as "exim-". This error state is now caught, and the installation is
4373 38. An argument was missing from the function that creates an error message
4374 when Exim fails to connect to the socket for saslauthd authentication.
4375 This could cause Exim to crash, or give a corrupted message.
4377 39. Added isip, isip4, and isip6 to ${if conditions.
4379 40. The ACL variables $acl_xx are now saved with the message, and can be
4380 accessed later in routers, transports, and filters.
4382 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4383 strings in the file are not string-expanded.
4385 42. If a MAIL command specified a SIZE value that was too large to fit into an
4386 int variable, the check against message_size_limit failed. Such values are
4387 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4388 day this will have to be increased, but I don't think I want to be around
4389 when emails are that large.
4396 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4397 iconv() is not standard in FreeBSD.
4399 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4400 IPv6 enabled. The observed symptom was a segmentation fault on return from
4401 the function os_common_find_running_interfaces() in src/os.c.
4403 3. In the check_special_case() function in daemon.c I had used "errno" as an
4404 argument name, which causes warnings on some systems. This was basically a
4405 typo, since it was named "eno" in the comments!
4407 4. The code that waits for the clock to tick (at a resolution of some fraction
4408 of a second) so as to ensure message-id uniqueness was always waiting for
4409 at least one whole tick, when it could have waited for less. [This is
4410 almost certainly not relevant at current processor speeds, where it is
4411 unlikely to ever wait at all. But we try to future-proof.]
4413 5. The function that sleeps for a time interval that includes fractions of a
4414 second contained a race. It did not block SIGALRM between setting the
4415 timer, and suspending (a couple of lines later). If the interval was short
4416 and the sigsuspend() was delayed until after it had expired, the suspension
4417 never ended. On busy systems this could lead to processes getting stuck for
4420 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4421 process, before it forks any delivery processes. The open lookup caching
4422 mechanism meant that the open file or database connection was passed into
4423 the delivery process. The problem was that delivery processes always tidy
4424 up cached lookup data. This could cause a problem for the next delivery
4425 process started by the queue runner, because the external queue runner
4426 process does not know about the closure. So the next delivery process
4427 still has data in the lookup cache. In the case of a file lookup, there was
4428 no problem because closing a file descriptor in a subprocess doesn't affect
4429 the parent. However, if the lookup was caching a connection to a database,
4430 the connection was closed, and the second delivery process was likely to
4431 see errors such as "PGSQL: query failed: server closed the connection
4432 unexpectedly". The problem has been fixed by closing all cached lookups
4433 in a queue runner before running a delivery process.
4435 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4436 seem to have the "const" qualifier which it has on other OS. I've
4439 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4440 *of the same type* (client or server) with the same public name that an
4441 error should be diagnosed.
4443 9. When Exim looked up a host name for an IP address, but failed to find the
4444 original IP address when looking up the host name (a safety check), it
4445 output the message "<ip address> does not match any IP for NULL", which was
4446 confusing, to say the least. The bug was that the host name should have
4447 appeared instead of "NULL".
4449 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4450 user that is built into the binary, and the -C or -D options is used, root
4451 privilege is dropped before the configuration file is read. In addition,
4452 logging is switched to stderr instead of the normal log files. If the
4453 configuration then re-defines the Exim user, the unprivileged environment
4454 is probably not what is expected, so Exim logs a panic warning message (but
4457 However, if deliver_drop_privilege is set, the unprivileged state may well
4458 be exactly what is intended, so the warning has been cut out in that case,
4459 and Exim is allowed to try to write to its normal log files.
4465 1. smtp_return_error_details was not giving details for temporary sender
4466 or receiver verification errors.
4468 2. Diagnose a configuration error if two authenticators have the same public
4471 3. Exim used not to create the message log file for a message until the first
4472 delivery attempt. This could be confusing when incoming messages were held
4473 for policy or load reasons. The message log file is now created at the time
4474 the message is received, and an initial "Received" line is written to it.
4476 4. The automatically generated man page for command line options had a minor
4477 bug that caused no ill effects; however, a more serious problem was that
4478 the procedure for building the man page automatically didn't always
4479 operate. Consequently, release 4.20 contains an out-of-date version. This
4480 shouldn't happen again.
4482 5. When building Exim with embedded Perl support, the script that builds the
4483 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4484 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4486 6. The freeze_tell option was not being used for messages that were frozen on
4487 arrival, either by an ACL or by local_scan().
4489 7. Added the smtp_incomplete_transaction log selector.
4491 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4492 was accepting AUTH without a new EHLO.
4494 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4495 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4498 10. Logging of TCP/IP connections (when configured) now happens in the main
4499 daemon process instead of the child process, so that the TCP/IP connection
4500 count is more accurate (but it can never be perfect).
4502 11. The use of "drop" in a nested ACL was not being handled correctly in the
4503 outer ACL. Now, if condition failure induced by the nested "drop" causes
4504 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4505 or "require"), the connection is dropped.
4507 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4508 that yield "discard" can now be used with an "accept" or a "discard" verb,
4509 but an error is generated for any others (because I can't see a useful way
4510 to define what should happen).
4512 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4513 are now processed in the same way as lines in the Exim configuration file.
4514 In particular, continuation lines are supported.
4516 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4518 15. Added -ti meaning -t -i.
4520 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4521 domains, and warn by logging if others are found.
4523 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4524 returned from GIFCONF when Exim is trying to find the list of interfaces on
4525 a host. The code in os.c has been modified to copy each ifreq to an aligned
4526 structure in all cases.
4528 Also, in some cases, the returned ifreq's were being copied to a 'struct
4529 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4530 means the last couple of bytes of an IPv6 address could be chopped if the
4531 ifreq contained only a normal sockaddr (14 bytes storage).
4533 18. Named domain lists were not supported in the hosts_treat_as_local option.
4534 An entry such as +xxxx was not recognized, and was treated as a literal
4537 19. Ensure that header lines added by a DATA ACL are included in the reject log
4538 if the ACL subsequently rejects the message.
4540 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4541 MD5 (which is deprecated).
4543 21. When testing a filter file using -bf, Exim was writing a message when it
4544 took the sender from a "From " line in the message, but it was not doing so
4545 when it took $return_path from a Return-Path: header line. It now does.
4547 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4548 with a valid header line field name (a series of printing characters
4549 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4551 23. Changed "disc" in the source to "disk" to conform to the documentation and
4552 the book and for uniformity.
4554 24. Ignore Sendmail's -Ooption=value command line item.
4556 25. When execve() failed while trying to run a command in a pipe transport,
4557 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4558 could be confused with a return value of 69 from the command itself. This
4559 has been changed to 127, the value the shell returns if it is asked to run
4560 a non-existent command. The wording for the related log line suggests a
4561 non-existent command as the problem.
4563 26. If received_header_text expands to an empty string, do not add a Received:
4564 header line to the message. (Well, it adds a token one on the spool, but
4565 marks it "old" so that it doesn't get used or transmitted.)
4567 27. Installed eximstats 1.28 (addition of -nt option).
4569 28. There was no check for failure on the call to getsockname() in the daemon
4570 code. This can fail if there is a shortage of resources on the system, with
4571 ENOMEM, for example. A temporary error is now given on failure.
4573 29. Contrary to the C standard, it seems that in some environments, the
4574 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4575 program. Exim now does this explicitly; it affects the formatting of
4576 timestamps using strftime().
4578 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4579 complaints. I've now initialized all the variables, to avoid this.
4581 32. Header lines added by a system filter were not being "seen" during
4582 transport-time rewrites.
4584 33. The info_callback() function passed to OpenSSL is set up with type void
4585 (*)(SSL *, int, int), as described somewhere. However, when calling the
4586 function (actually a macro) that sets it up, the type void(*)() is
4587 expected. I've put in a cast to prevent warnings from picky compilers.
4589 34. If a DNS black list lookup found a CNAME record, but there were no A
4590 records associated with the domain it pointed at, Exim crashed.
4592 35. If a DNS black list lookup returned more than one A record, Exim ignored
4593 all but the first. It now scans all returned addresses if a particular IP
4594 value is being sought. In this situation, the contents of the
4595 $dnslist_value variable are a list of all the addresses, separated by a
4598 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4599 to accept a host name and all its aliases if the forward lookup for any of
4600 them yielded the IP address of the incoming connection. Now it accepts only
4601 those names whose forward lookup yields the correct IP address. Any other
4602 names are discarded. This closes a loophole whereby a rogue DNS
4603 administrator could create reverse DNS records to break through a
4604 wildcarded host restriction in an ACL.
4606 37. If a user filter or a system filter that ran in a subprocess used any of
4607 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4608 the wrong values were passed to the pipe command ($thisaddress had the
4609 value of $0, $0 had the value of $1, etc). This bug was introduced by
4610 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4612 38. Improved the line breaking for long SMTP error messages from ACLs.
4613 Previously, if there was no break point between 40 and 75 characters, Exim
4614 left the rest of the message alone. Two changes have been made: (a) I've
4615 reduced the minimum length to 35 characters; (b) if it can't find a break
4616 point between 35 and 75 characters, it looks ahead and uses the first one
4617 that it finds. This may give the occasional overlong line, but at least the
4618 remaining text gets split now.
4620 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4621 file descriptors might be low, and that setting 1000 would always raise it.
4622 It turns out that in some environments, the limit is already over 1000 and
4623 that lowering it causes trouble. So now Exim takes care not to decrease it.
4625 40. When delivering a message, the value of $return_path is set to $sender_
4626 address at the start of routing (routers may change the value). By an
4627 oversight, this default was not being set up when an address was tested by
4628 -bt or -bv, which affected the outcome if any router or filter referred to
4631 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4632 log only when "message" or "log_message" are set. However, if one of the
4633 conditions was an address verification, or a call to a nested ACL, the
4634 messages generated by the underlying test were being passed through. This
4635 no longer happens. The underlying message is available in $acl_verify_
4636 message for both "message" and "log_message" expansions, so it can be
4637 passed through if needed.
4639 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4640 new expansion $bh_ to give the encoded byte string without charset
4641 translation. Translation happens only if iconv() is available; HAVE_ICONV
4642 indicates this at build time. HEADERS_CHARSET gives the charset to
4643 translate to; headers_charset can change it in the configuration, and
4644 "headers charset" can change it in an individual filter file.
4646 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4647 that creates RFC 2047 encoded "words" labels them as that charset instead
4648 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4649 expansion operator; (ii) when Exim creates a From: line for a local
4650 message; (iii) when a header line is rewritten to include a "phrase" part.
4652 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4653 buggy, causing it to skip the first lines of messages whose message ID
4654 ended in 'D'. This would not have bitten before Exim release 4.14, because
4655 message IDs were unlikely to end in 'D' before then. The effect was to have
4656 incorrect size information for certain domains.
4658 45. #include "config.h" was missing at the start of the crypt16.c module. This
4659 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4662 46. If there was a timeout during a "random" callout check, Exim treated it as
4663 a failure of the random address, and carried on sending RSET and the real
4664 address. If the delay was just some slowness somewhere, the response to the
4665 original RCPT would be taken as a response to RSET and so on, causing
4666 mayhem of various kinds.
4668 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4669 when I implemented it. It didn't allow for the fact that some option values
4670 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4671 the right test for positive values.
4673 48. Domain names in DNS records are case-independent. Exim always looks them up
4674 in lower case. Some resolvers return domain names in exactly the case they
4675 appear in the zone file, that is, they may contain uppercase letters. Not
4676 all resolvers do this - some return always lower case. Exim was treating a
4677 change of case by a resolver as a change of domain, similar to a widening
4678 of a domain abbreviation. This triggered its re-routing code and so it was
4679 trying to route what was effectively the same domain again. This normally
4680 caused routing to fail (because the router wouldn't handle the domain
4681 twice). Now Exim checks for this case specially, and just changes the
4682 casing of the domain that it ultimately uses when it transmits the message
4685 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4688 50. If a filter generated a file delivery with a non-absolute name (possible if
4689 no home directory exists for the router), the forbid_file option was not
4692 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4693 the existing equality matching.
4695 52. Exim was using ints instead of ino_t variables in some places where it was
4696 dealing with inode numbers.
4698 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4699 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4700 called TMPDIR, and if it finds it is different, it changes its value.
4702 54. The smtp_printf() function is now made available to local_scan() so
4703 additional output lines can be written before returning. There is also an
4704 smtp_fflush() function to enable the detection of a dropped connection.
4705 The variables smtp_input and smtp_batched_input are exported to
4708 55. Changed the default runtime configuration: the message "Unknown user"
4709 has been removed from the ACL, and instead placed on the localuser router,
4710 using the cannot_route_message feature. This means that any verification
4711 failures that generate their own messages won't get overridden. Similarly,
4712 the "Unrouteable address" message that was in the ACL for unverifiable
4713 relay addresses has also been removed.
4715 56. Added hosts_avoid_esmtp to the smtp transport.
4717 57. The exicyclog script was not checking for the esoteric option
4718 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4719 will work only if exicyclog is run under the appropriate euid.
4721 58. Following a discussion on the list, the rules by which Exim recognises line
4722 endings on incoming messages have been changed. The -dropcr and drop_cr
4723 options are now no-ops, retained only for backwards compatibility. The
4724 following line terminators are recognized: LF CRLF CR. However, special
4725 processing applies to CR:
4727 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4728 nor a local message in the state where . is a terminator.
4730 (ii) If a bare CR is encountered in a header line, an extra space is added
4731 after the line terminator so as not to end the header. The reasoning
4732 behind this is that bare CRs in header lines are most likely either
4733 to be mistakes, or people trying to play silly games.
4735 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4736 was being incorrectly given as 18 bytes larger than it should have been.
4737 This is a VOB (very old bug).
4739 60. This may never have affected anything current, but just in case it has:
4740 When the local host is found other than at the start of a list of hosts,
4741 the local host, those with the same MX, and any that follow, are discarded.
4742 When the list in question was part of a longer list of hosts, the following
4743 hosts (not currently being processed) were also being discarded. This no
4744 longer happens. I'm not sure if this situation could ever has previously
4747 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4750 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4751 header. This is something that is recommended in a new Internet Draft, and
4752 is something that is documented as being done by Sendmail. There are two
4753 possible values. For messages generated by the autoreply transport, Exim
4756 Auto-Submitted: auto-replied
4758 whereas for all other generated messages (e.g. bounces) it adds
4760 Auto-Submitted: auto-generated
4762 63. The "personal" condition in filters now includes a test for the
4763 Auto-Submitted: header. If it contains the string "auto-" the message it
4764 not considered personal.
4766 64. Added rcpt_include_affixes as a generic transport option.
4768 65. Added queue_only_override (default true).
4770 66. Added the syslog_duplication option.
4772 67. If what should have been the first header line of a message consisted of
4773 a space followed by a colon, Exim was mis-interpreting it as a header line.
4774 It isn't of course - it is syntactically invalid and should therefore be
4775 treated as the start of the message body. The misbehaviour could have
4776 caused a number of strange effects, including loss of data in subsequent
4777 header lines, and spool format errors.
4779 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4780 client host had authenticated. This control can now be exercised by an ACL
4781 for more flexibility.
4783 69. By default, callouts do not happen when testing with -bh. There is now a
4784 variant, -bhc, which does actually run the callout code, including
4785 consulting and updating the callout cache.
4787 70. Added support for saslauthd authentication, courtesy of Alexander
4790 71. If statvfs() failed on the spool or log directories while checking their
4791 size for availability, Exim confusingly gave the error "space shortage".
4792 Furthermore, in debugging mode it crashed with a floating point exception.
4793 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4794 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4795 serious problem, Exim now writes to the main and panic logs when this
4796 happens, with details of the failure. It then refuses to accept the
4797 incoming message, giving the message "spool directory problem" or "log
4798 directory problem" with a 421 code for SMTP messages.
4800 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4801 0, 1, and 2 exist, because some OS complain for execs without them (see
4802 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4803 descriptors. However, the code omitted to check that the open succeeded,
4804 causing mysterious errors if for some reason the permissions on /dev/null
4805 got screwed. Now Exim writes a message to the main and panic logs, and
4806 bombs out if it can't open /dev/null.
4808 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4809 interact so that it is all more flexible. It is supposed to remain
4810 backwards compatible. Also added extra_local_interfaces.
4812 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4813 to bomb out with an assertion failure - to the client this appears as a
4814 connection drop. This problem occurs in the part of the code that was taken
4815 from the Samba project. Fortunately, the assertion is in a very simple
4816 function, so I have fixed this by reproducing the function inline in the
4817 one place where it is called, and arranging for authentication to fail
4818 instead of killing the process with assert().
4820 75. The SPA client code was not working when the server requested OEM rather
4821 than Unicode encoding.
4823 76. Added code to make require_files with a specific uid setting more usable in
4824 the case where statting the file as root fails - usually a non-root-mounted
4825 NFS file system. When this happens and the failure is EACCES, Exim now
4826 forks a subprocess and does the per-uid checking as the relevant uid.
4828 77. Added process_log_path.
4830 78. If log_file_path was not explicitly set, a setting of check_log_space or
4831 check_log_inodes was ignored.
4833 79. If a space check for the spool or log partitions fails, the incident is now
4834 logged. Of course, in the latter case the data may get lost...
4836 80. Added the %p formatting code to string_format() so that it can be used to
4837 print addresses in debug_print(). Adjusted all the address printing in the
4838 debugging in store.c to use %p rather than %d.
4840 81. There was a concern that a line of code in smtp_in.c could overflow a
4841 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4842 initially expressed, the concern was not well-founded, because trailing
4843 spaces are removed early. However, if the trailing spaces were followed by
4844 a NULL, they did not get removed, so the overflow was possible. Two fixes
4847 (a) I re-wrote the offending code in a cleaner fashion.
4848 (b) If an incoming SMTP command contains a NULL character, it is rejected
4851 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4852 runs initgroups(), so that if the Exim user is in any additional groups,
4853 they will be used during message reception.
4859 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob