1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.537 2008/01/17 12:56:13 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
10 When building exim an external PCRE library is now needed -
11 PCRE is a system library on the majority of modern systems.
12 See entry on PCRE_LIBS in EDITME file.
17 TK/01 Add preliminary DKIM support. Currently requires a forked version of
18 ALT-N's libdkim that I have put here:
19 http://duncanthrax.net/exim-experimental/
21 Note to Michael Haardt: I had to rename some vars in sieve.c. They
22 were called 'true' and it seems that C99 defines that as a reserved
23 keyword to be used with 'bool' variable types. That means you could
24 not include C99-style headers which use bools without triggering
25 build errors in sieve.c.
27 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
28 as mailq or other aliases. Changed the --help handling significantly
29 to do whats expected. exim_usage() emits usage/help information.
31 SC/01 Added the -bylocaldomain option to eximstats.
33 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
35 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
37 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
43 PH/01 Another patch from the Sieve maintainer.
45 PH/02 When an IPv6 address is converted to a string for single-key lookup
46 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
47 dots are used instead of colons so that keys in lsearch files need not
48 contain colons. This was done some time before quoting was made available
49 in lsearch files. However, iplsearch files do require colons in IPv6 keys
50 (notated using the quote facility) so as to distinguish them from IPv4
51 keys. This meant that lookups for IP addresses in host lists did not work
52 for iplsearch lookups.
54 This has been fixed by arranging for IPv6 addresses to be expressed with
55 colons if the lookup type is iplsearch. This is not incompatible, because
56 previously such lookups could never work.
58 The situation is now rather anomolous, since one *can* have colons in
59 ordinary lsearch keys. However, making the change in all cases is
60 incompatible and would probably break a number of configurations.
62 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
65 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
66 conversion specification without a maximum field width, thereby enabling
67 a rogue spamd server to cause a buffer overflow. While nobody in their
68 right mind would setup Exim to query an untrusted spamd server, an
69 attacker that gains access to a server running spamd could potentially
70 exploit this vulnerability to run arbitrary code as the Exim user.
72 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
73 $primary_hostname instead of what libspf2 thinks the hosts name is.
75 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
76 a directory entry by the name of the lookup key. Previously, if a
77 symlink pointed to a non-existing file or a file in a directory that
78 Exim lacked permissions to read, a lookup for a key matching that
79 symlink would fail. Now it is enough that a matching directory entry
80 exists, symlink or not. (Bugzilla 503.)
82 PH/03 The body_linecount and body_zerocount variables are now exported in the
85 PH/04 Added the $dnslist_matched variable.
87 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
88 This means they are set thereafter only if the connection becomes
91 PH/06 Added the client_condition to authenticators so that some can be skipped
92 by clients under certain conditions.
94 PH/07 The error message for a badly-placed control=no_multiline_responses left
95 "_responses" off the end of the name.
97 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
99 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
100 (without spaces) instead of just copying the configuration text.
102 PH/10 Added the /noupdate option to the ratelimit ACL condition.
104 PH/11 Added $max_received_linelength.
106 PH/12 Added +ignore_defer and +include_defer to host lists.
108 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
109 way in which PCRE > 7.0 is built.
111 PH/14 Implemented queue_only_load_latch.
113 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
114 MAIL command. The effect was to mangle the value on 64-bit systems.
116 PH/16 Another patch from the Sieve maintainer.
118 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
120 PH/18 If a system quota error occurred while trying to create the file for
121 a maildir delivery, the message "Mailbox is full" was not appended to the
122 bounce if the delivery eventually timed out. Change 4.67/27 below applied
123 only to a quota excession during the actual writing of the file.
125 PH/19 It seems that peer DN values may contain newlines (and other non-printing
126 characters?) which causes problems in log lines. The DN values are now
127 passed through string_printing() before being added to log lines.
129 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
130 and InterBase are left for another time.)
132 PH/21 Added message_body_newlines option.
134 PH/22 Guard against possible overflow in moan_check_errorcopy().
136 PH/23 POSIX allows open() to be a macro; guard against that.
138 PH/24 If the recipient of an error message contained an @ in the local part
139 (suitably quoted, of course), incorrect values were put in $domain and
140 $local_part during the evaluation of errors_copy.
146 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
147 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
150 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
151 issue a MAIL command.
153 PH/02 In an ACL statement such as
155 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
157 if a client was not listed at all, or was listed with a value other than
158 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
159 the condition was not true (as it should be), so access was not denied.
160 The bug was that the ! inversion was incorrectly passed on to the second
161 item. This has been fixed.
163 PH/03 Added additional dnslists conditions == and =& which are different from
164 = and & when the dns lookup returns more than one IP address.
166 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
167 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
169 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
170 FSYNC, which compiles an option called disable_fsync that allows for
171 bypassing fsync(). The documentation is heavily laced with warnings.
173 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
175 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
176 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
177 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
178 including adding "make clean"; (3) Added -fPIC when compiling the test
179 dynamically loaded module, to get rid of a warning.
181 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
182 message fails, move_frozen_messages = true and ignore_bounce_errors_after
183 = 0s. The bug is otherwise harmless.
185 PH/07 There was a bug in the dovecot authenticator such that the value of
186 $auth1 could be overwritten, and so not correctly preserved, after a
187 successful authentication. This usually meant that the value preserved by
188 the server_setid option was incorrect.
190 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
192 PH/09 Installed PCRE release 7.0.
194 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
195 run for batched SMTP input. It is now run at the start of every message
196 in the batch. While fixing this I discovered that the process information
197 (output by running exiwhat) was not always getting set for -bs and -bS
198 input. This is fixed, and it now also says "batched" for BSMTP.
200 PH/11 Added control=no_pipelining.
202 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
203 patch, slightly modified), and move the expansion of helo_data till after
204 the connection is made in the smtp transport (so it can use these
207 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
209 PH/14 Added log_selector = +pid.
211 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
213 PH/16 Add ${if forany and ${if forall.
215 PH/17 Added dsn_from option to vary the From: line in DSNs.
217 PH/18 Flush SMTP output before performing a callout, unless control =
218 no_callout_flush is set.
220 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
221 was true (the default) a successful delivery failed to delete the retry
222 item, thus causing premature timeout of the address. The bug is now
225 PH/20 Added hosts_avoid_pipelining to the smtp transport.
227 PH/21 Long custom messages for fakedefer and fakereject are now split up
228 into multiline reponses in the same way that messages for "deny" and
229 other ACL rejections are.
231 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
232 with slight modification.
234 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
235 draft, changing the syntax and factoring some duplicate code".
237 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
238 for deliveries of the second and subsequent messages over the same SMTP
241 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
242 ${reduce, with only minor "tidies".
244 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
246 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
247 expansion side effects.
249 PH/27 When a message times out after an over-quota error from an Exim-imposed
250 quota, the bounce message says "mailbox is full". This message was not
251 being given when it was a system quota that was exceeded. It now should
254 MH/03 Made $recipients available in local_scan(). local_scan() already has
255 better access to the recipient list through recipients_list[], but
256 $recipients can be useful in postmaster-provided expansion strings.
258 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
259 in the case of a MAIL command with additional options following the
260 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
261 were accidentally chopped off.
263 PH/29 SMTP synchronization checks are implemented when a command is read -
264 there is a check that no more input is waiting when there shouldn't be
265 any. However, for some commands, a delay in an ACL can mean that it is
266 some time before the response is written. In this time, more input might
267 arrive, invalidly. So now there are extra checks after an ACL has run for
268 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
269 pipelining has not been advertised.
271 PH/30 MH's patch to allow iscntrl() characters to be list separators.
273 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
274 returned in the SMTP response when smtp_return_error_details was false.
277 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
278 instead of the C I/O that was originally supplied, because problems were
281 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
282 Exim which did not show up earlier: it was assuming that a call to
283 SSL_CTX_set_info_callback() might give an error value. In fact, there is
284 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
285 was a macro that became an assignment, so it seemed to work. This has
286 changed to a proper function call with a void return, hence the compile
287 error. Exim's code has been fixed.
289 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
292 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
294 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
296 PH/37 If a message is not accepted after it has had an id assigned (e.g.
297 because it turns out to be too big or there is a timeout) there is no
298 "Completed" line in the log. When some messages of this type were
299 selected by exigrep, they were listed as "not completed". Others were
300 picked up by some special patterns. I have improved the selection
301 criteria to be more general.
303 PH/38 The host_find_failed option in the manualroute router can now be set
304 to "ignore", to completely ignore a host whose IP address cannot be
305 found. If all hosts are ignored, the behaviour is controlled by the new
306 host_all_ignored option.
308 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
309 homing or because of multiple MX records with /mx) generated more than
310 one IP address, and the following item turned out to be the local host,
311 all the secondary addresses of the first item were incorrectly removed
312 from the list, along with the local host and any following hosts (which
313 is what is supposed to happen).
315 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
316 whoever called Exim into the -H file. In the case of the daemon it was
317 behaving confusingly. When first started, it used values for whoever
318 started the daemon, but after a SIGHUP it used the Exim user (because it
319 calls itself on a restart). I have changed the code so that it now always
322 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
323 message are rejected with the same error (e.g. no authentication or bad
324 sender address), and a DATA command is nevertheless sent (as can happen
325 with PIPELINING or a stupid MUA), the error message that was given to the
326 RCPT commands is included in the rejection of the DATA command. This is
327 intended to be helpful for MUAs that show only the final error to their
330 PH/42 Another patch from the Sieve maintainer.
332 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
333 Eximstats - Fixed some broken HTML links and added missing column headers
335 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
336 columns for Rejects, Temp Rejects, Ham, and Spam rows.
338 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
340 PH/43 Yet another patch from the Sieve maintainer.
342 PH/44 I found a way to check for a TCP/IP connection going away before sending
343 the response to the final '.' that terminates a message, but only in the
344 case where the client has not sent further data following the '.'
345 (unfortunately, this is allowed). However, in many cases there won't be
346 any further data because there won't be any more messages to send. A call
347 to select() can be used: if it shows that the input is "ready", there is
348 either input waiting, or the socket has been closed. An attempt to read
349 the next input character can distinguish the two cases. Previously, Exim
350 would have sent an OK response which the client would never have see.
351 This could lead to message repetition. This fix should cure that, at
352 least in a lot of common cases.
354 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
355 advertised in response to EHLO.
361 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
362 fixed by 4.65/MH/01 (is this a record?) are fixed:
364 (i) An empty string was always treated as zero by the numeric comparison
365 operators. This behaviour has been restored.
367 (ii) It is documented that the numeric comparison operators always treat
368 their arguments as decimal numbers. This was broken in that numbers
369 starting with 0 were being interpreted as octal.
371 While fixing these problems I realized that there was another issue that
372 hadn't been noticed. Values of message_size_limit (both the global option
373 and the transport option) were treated as octal if they started with 0.
374 The documentation was vague. These values are now always treated as
375 decimal, and I will make that clear in the documentation.
381 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
382 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
385 MH/01 Don't check that the operands of numeric comparison operators are
386 integers when their expansion is in "skipping" mode (fixes bug
387 introduced by 4.64-PH/07).
389 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
390 child addresses, Exim now panics and dies. Previously, because the count
391 is held in a short int, deliveries were likely to be lost. As such a
392 large number of recipients for a single message is ridiculous
393 (performance will be very, very poor), I have chosen to impose a limit
394 rather than extend the field.
400 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
401 leftover -K file (the existence of which was triggered by #402).
402 While we were at it, introduced process PID as part of the -K
403 filename. This should rule out race conditions when creating
406 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
407 processing considerably. Previous code took too long for large mails,
408 triggering a timeout which in turn triggers #401.
410 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
411 in the DK code in transports.c. sendfile() is not really portable,
412 hence the _LINUX specificness.
414 TF/01 In the add_headers option to the mail command in an Exim filter,
415 there was a bug that Exim would claim a syntax error in any
416 header after the first one which had an odd number of characters
419 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
420 callout verification, Exim cached a "reject" for the entire domain. This
421 is correct for most verifications, but it is not correct for a recipient
422 verification with use_sender or use_postmaster set, because in that case
423 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
424 case of MAIL FROM:<> rejection from other early rejections (e.g.
425 rejection of HELO). When verifying a recipient using a non-null MAIL
426 address, the cache is ignored if it shows MAIL FROM:<> rejection.
427 Whatever the result of the callout, the value of the domain cache is
428 left unchanged (for any other kind of callout, getting as far as trying
429 RCPT means that the domain itself is ok).
431 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
434 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
435 manifest itself as EPIPE rather than ECONNECT. When tidying away a
436 session, the daemon ignores ECONNECT errors and logs others; it now
437 ignores EPIPE as well.
439 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
440 (quoted-printable decoding).
442 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
443 later the small subsequent patch to fix an introduced bug.
445 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
447 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
449 PH/08 An error is now given if message_size_limit is specified negative.
451 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
452 to be given (somewhat) arbitrary names.
454 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
457 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
458 miscellaneous code fixes
460 PH/10 Added the log_reject_target ACL modifier to specify where to log
463 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
464 hostname. This is wrong, because it relates to the incoming message (and
465 probably the interface on which it is arriving) and not to the outgoing
466 callout (which could be using a different interface). This has been
467 changed to use the value of the helo_data option from the smtp transport
468 instead - this is what is used when a message is actually being sent. If
469 there is no remote transport (possible with a router that sets up host
470 addresses), $smtp_active_hostname is used.
472 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
473 tweaks were necessary in order to get it to work (see also 21 below):
474 (a) The code assumed that strncpy() returns a negative number on buffer
475 overflow, which isn't the case. Replaced with Exim's string_format()
477 (b) There were several signed/unsigned issues. I just did the minimum
478 hacking in of casts. There is scope for a larger refactoring.
479 (c) The code used strcasecmp() which is not a standard C function.
480 Replaced with Exim's strcmpic() function.
481 (d) The code set only $1; it now sets $auth1 as well.
482 (e) A simple test gave the error "authentication client didn't specify
483 service in request". It would seem that Dovecot has changed its
484 interface. Fortunately there's a specification; I followed it and
485 changed what the client sends and it appears to be working now.
487 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
490 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
491 address A is aliased to B and C, where B exists and C does not. Without
492 -v the output is "A verified" because verification stops after a
493 successful redirection if more than one address is generated. However,
494 with -v the child addresses are also verified. Exim was outputting "A
495 failed to verify" and then showing the successful verification for C,
496 with its parentage. It now outputs "B failed to verify", showing B's
497 parentage before showing the successful verification of C.
499 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
500 look up a TXT record in a specific list after matching in a combined
503 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
504 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
505 they consult the DNS. I had assumed they would set it the way they
506 wanted; and indeed my experiments on Linux seem to show that in some
507 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
508 To be on the safe side, however, I have now made the interface to
509 host_find_byname() similar to host_find_bydns(), with an argument
510 containing the DNS resolver options. The host_find_byname() function now
511 sets these options at its start, just as host_find_bydns() does. The smtp
512 transport options dns_qualify_single and dns_search_parents are passed to
513 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
514 of host_find_byname() use the default settings of RES_DEFNAMES
515 (qualify_single) but not RES_DNSRCH (search_parents).
517 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
518 spool_read_header() do less string testing, by means of a preliminary
519 switch on the second character of optional "-foo" lines. (This is
520 overdue, caused by the large number of possibilities that now exist.
521 Originally there were few.) While I was there, I also converted the
522 str(n)cmp tests so they don't re-test the leading "-" and the first
523 character, in the hope this might squeeze out yet more improvement.
525 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
526 flag allowing group syntax was set by the header_syntax check but not
527 turned off, possible causing trouble later; (2) The flag was not being
528 set at all for the header_verify test, causing "group"-style headers to
529 be rejected. I have now set it in this case, and also caused header_
530 verify to ignore an empty address taken from a group. While doing this, I
531 came across some other cases where the code for allowing group syntax
532 while scanning a header line wasn't quite right (mostly, not resetting
533 the flag correctly in the right place). These bugs could have caused
534 trouble for malformed header lines. I hope it is now all correct.
536 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
537 with the "reply" argument non-NULL. The code, however (which originally
538 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
539 but it didn't always do it. This confused somebody who was copying the
540 code for some other use. I have removed all the tests.
542 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
543 feature that was used to support insecure browsers during the U.S. crypto
544 embargo. It requires special client support, and Exim is probably the
545 only MTA that supported it -- and would never use it because real RSA is
546 always available. This code has been removed, because it had the bad
547 effect of slowing Exim down by computing (never used) parameters for the
548 RSA_EXPORT functionality.
550 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
551 authenticator to fail if there's a tab character in the incoming data
552 (there should never be unless someone is messing about, as it's supposed
553 to be base64-encoded). Also added, on Timo's advice, the "secured" option
554 if the connection is using TLS or if the remote IP is the same as the
555 local IP, and the "valid-client-cert option" if a client certificate has
558 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
559 authenticators. This can be used for authorization after authentication
560 succeeds. (In the case of plaintext, it servers for both authentication
563 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
564 if any retry times were supplied.
566 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
567 connection, where there is no client IP address to check. In this
568 situation, the verify now always succeeds.
570 PH/25 Applied John Jetmore's -Mset patch.
572 PH/26 Added -bem to be like -Mset, but loading a message from a file.
574 PH/27 In a string expansion for a processed (not raw) header when multiple
575 headers of the same name were present, leading whitespace was being
576 removed from all of them, but trailing whitespace was being removed only
577 from the last one. Now trailing whitespace is removed from each header
578 before concatenation. Completely empty headers in a concatenation (as
581 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
582 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
584 PH/29 [Removed. This was a change that I later backed out, and forgot to
585 correct the ChangeLog entry (that I had efficiently created) before
586 committing the later change.]
588 PH/30 Exim was sometimes attempting to deliver messages that had suffered
589 address errors (4xx response to RCPT) over the same connection as other
590 messages routed to the same hosts. Such deliveries are always "forced",
591 so retry times are not inspected. This resulted in far too many retries
592 for the affected addresses. The effect occurred only when there were more
593 hosts than the hosts_max_try setting in the smtp transport when it had
594 the 4xx errors. Those hosts that it had tried were not added to the list
595 of hosts for which the message was waiting, so if all were tried, there
596 was no problem. Two fixes have been applied:
598 (i) If there are any address or message errors in an SMTP delivery, none
599 of the hosts (tried or untried) are now added to the list of hosts
600 for which the message is waiting, so the message should not be a
601 candidate for sending over the same connection that was used for a
602 successful delivery of some other message. This seems entirely
603 reasonable: after all the message is NOT "waiting for some host".
604 This is so "obvious" that I'm not sure why it wasn't done
605 previously. Hope I haven't missed anything, but it can't do any
606 harm, as the worst effect is to miss an optimization.
608 (ii) If, despite (i), such a delivery is accidentally attempted, the
609 routing retry time is respected, so at least it doesn't keep
610 hammering the server.
612 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
613 in ${readsocket because some servers need this prod.
615 PH/32 Added some extra debug output when updating a wait-xxx database.
617 PH/33 The hint "could be header name not terminated by colon", which has been
618 given for certain expansion errors for a long time, was not being given
619 for the ${if def:h_colon_omitted{... case.
621 PH/34 The spec says: "With one important exception, whenever a domain list is
622 being scanned, $domain contains the subject domain." There was at least
623 one case where this was not true.
625 PH/35 The error "getsockname() failed: connection reset by peer" was being
626 written to the panic log as well as the main log, but it isn't really
627 panic-worthy as it just means the connection died rather early on. I have
628 removed the panic log writing for the ECONNRESET error when getsockname()
631 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
632 runs only) independently of the message's sender address. This meant
633 that, if the 4xx error was in fact related to the sender, a different
634 message to the same recipient with a different sender could confuse
635 things. In particualar, this can happen when sending to a greylisting
636 server, but other circumstances could also provoke similar problems.
637 I have changed the default so that the retry time for these errors is now
638 based a combination of the sender and recipient addresses. This change
639 can be overridden by setting address_retry_include_sender=false in the
642 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
643 remote server are returned as part of bounce messages. This was not
644 happening for LMTP over a pipe (the lmtp transport), but now it is the
645 same for both kinds of LMTP.
647 PH/38 Despite being documented as not happening, Exim was rewriting addresses
648 in header lines that were in fact CNAMEs. This is no longer the case.
650 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
651 and queue runs started by the daemon processed all messages. This has
652 been fixed so that -R and -S can now usefully be given with -q<time>.
654 PH/40 Import PCRE release 6.7 (fixes some bugs).
656 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
658 PH/42 Give an error if -q is specified more than once.
660 PH/43 Renamed the variables $interface_address and $interface_port as
661 $received_ip_address and $received_port, to make it clear that these
662 values apply to message reception, and not to the outgoing interface when
663 a message is delivered. (The old names remain recognized, of course.)
665 PH/44 There was no timeout on the connect() call when using a Unix domain
666 socket in the ${readsocket expansion. There now is.
668 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
669 be meaningful with "accept".
671 SC/01 Eximstats V1.43
672 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
674 SC/02 Eximstats V1.44
675 Use a glob alias rather than an array ref in the generated
676 parser. This improves both readability and performance.
678 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
679 Collect SpamAssassin and rejection statistics.
680 Don't display local sender or destination tables unless
681 there is data to show.
682 Added average volumes into the top table text output.
684 SC/04 Eximstats V1.46
685 Collect data on the number of addresses (recipients)
686 as well as the number of messages.
688 SC/05 Eximstats V1.47
689 Added 'Message too big' to the list of mail rejection
690 reasons (thanks to Marco Gaiarin).
692 SC/06 Eximstats V1.48
693 Mainlog lines which have GMT offsets and are too short to
694 have a flag are now skipped.
696 SC/07 Eximstats V1.49 (Alain Williams)
697 Added the -emptyok flag.
699 SC/08 Eximstats V1.50
700 Fixes for obtaining the IP address from reject messages.
702 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
703 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
704 whitesspace changes from 4.64-PH/27
706 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
709 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
710 are found, allow negative numbers in numeric criteria)
712 JJ/06 exipick.20061117.2, added new $message_body_missing variable
714 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
715 to match changes made in 4.64-PH/43
717 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
719 PH/47 Put in an explicit test for a DNS lookup of an address record where the
720 "domain" is actually an IP address, and force a failure. This locks out
721 those revolvers/nameservers that support "A-for-A" lookups, in
722 contravention of the specifications.
724 PH/48 When a host name was looked up from an IP address, and the subsequent
725 forward lookup of the name timed out, the host name was left in
726 $sender_host_name, contrary to the specification.
728 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
729 restricted to single-key lookups, Exim was not diagnosing an error if
730 * or *@ was used with a query-style lookup.
732 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
734 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
735 long ago, but noone interested enough thought of it. Let's just say that
736 the "1.1" means that there are some new functions that weren't there at
737 some point in the past.
739 PH/51 Error processing for expansion failure of helo_data from an smtp
740 transport during callout processing was broken.
742 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
743 tested/used via the -bh/-bhc/-bs options.
745 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
746 bug, fixed in subsequent PCRE releases).
748 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
749 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
751 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
757 SC/01 Use a glob alias rather than an array ref in eximstats generated
758 parser. This improves both readability and performance.
760 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
761 Don't display local sender or destination tables in eximstats unless
762 there is data to show.
763 Added average volumes into the eximstats top table text output.
765 SC/03 Collect data on the number of addresses (recipients) as well
766 as the number of messages in eximstats.
768 TF/01 Correct an error in the documentation for the redirect router. Exim
769 does (usually) call initgroups() when daemonizing.
771 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
772 with consistent privilege compared to when running as a daemon.
774 TF/03 Note in the spec that $authenticated_id is not set for local
775 submissions from trusted users.
777 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
778 Thanks to Dean Brooks <dean@iglou.com> for the patch.
780 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
781 by adding some example configuration directives to the default
782 configuration file. A little bit of work is required to uncomment the
783 directives and define how usernames and passwords are checked, but
784 there is now a framework to start from.
786 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
787 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
788 without this. I don't know how relevant this is to other LDAP libraries.
790 PH/02 Add the verb name to the "unknown ACL verb" error.
792 PH/03 Magnus Holmgren's patch for filter_prepend_home.
794 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
796 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
797 directory not expanded when it should be if an expanded home directory
798 was set for the address (which is overridden by the transport).
800 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
803 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
804 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
805 because it is too late at that time, and has no effect.
807 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
808 security issue with \' (bugzilla #107). I could not use the
809 PQescapeStringConn() function, because it needs a PGconn value as one of
812 PH/08 When testing addresses using -bt, indicate those final addresses that
813 are duplicates that would not cause an additional delivery. At least one
814 person was confused, thinking that -bt output corresponded to deliveries.
815 (Suppressing duplicates isn't a good idea as you lose the information
816 about possibly different redirections that led to the duplicates.)
818 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
819 systems where poll() doesn't work, in particular OS X.
821 PH/10 Added more information to debugging output for retry time not reached.
823 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
824 operations in malware.c.
826 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
829 PH/13 If write_rejectlog was set false when logging was sent to syslog with
830 syslog_duplication set false, log lines that would normally be written
831 both the the main log and to the reject log were not written to syslog at
834 PH/14 In the default configuration, change the use of "message" in ACL warn
835 statements to "add_header".
837 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
838 not followed by a command (e.g. "seen endif").
840 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
841 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
844 PH/17 Added extra conditions to the default value of delay_warning_condition
848 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
849 { match{$h_precedence:}{(?i)bulk|list|junk} } \
850 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
853 The Auto-Submitted: and various List- headers are standardised, whereas I
854 don't think Precedence: ever was.
856 PH/18 Refactored debugging code in route_finduser() to show more information,
857 in particular, the error code if getpwnam() issues one.
859 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
860 This is apparently needed in addition to the PH/07 change above to avoid
861 any possible encoding problems.
863 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
864 but not after initializing Perl.
866 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
867 output them only if debugging. By default they are written stderr,
868 apparently, which is not desirable.
870 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
873 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
876 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
878 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
879 authenticated or an ident call has been made. Suppress the default
880 values for $authenticated_id and $authenticated_sender (but permit -oMai
881 and -oMas) when testing with -bh.
883 PH/24 Re-jigged the order of the tests in the default configuration so that the
884 tests for valid domains and recipients precede the DNS black list and CSA
885 tests, on the grounds that those ones are more expensive.
887 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
888 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
889 This bug exists in every version of Exim that I still have, right back to
892 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
893 However, an attempt to turn on case-sensitivity in a regex key by
894 including (?-i) didn't work because the subject string was already
895 lowercased, and the effects were non-intuitive. It turns out that a
896 one-line patch can be used to allow (?-i) to work as expected.
902 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
903 other effects) broke the use of negated acl sub-conditions.
905 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
908 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
909 "Deny" causes Exim to reject the incoming connection with a 554 error.
910 Unfortunately, if there is a major crisis, such as a disk failure,
911 tcp-wrappers gives "deny", whereas what one would like would be some
912 kind of temporary error. A kludge has been added to help with this.
913 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
914 554 error is used if errno is still zero or contains ENOENT (which occurs
915 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
918 PH/03 Add -lutil to the default FreeBSD LIBS setting.
920 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
921 errors. Otherwise a message that provokes a temporary error (when other
922 messages do not) can cause a whole host to time out.
924 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
925 addresses were routed directly to files or pipes from a redirect router.
926 File deliveries just didn't batch; pipe deliveries might have suffered
929 PH/06 A failure to get a lock for a hints database would erroneously always say
930 "Failed to get write lock", even when it was really a read lock.
932 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
933 of 0600. This has been changed to use the value of the lockfile_mode
934 option (which defaults to 0600).
936 PH/08 Applied small patch from the Sieve maintainer.
938 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
939 folder from quota calculations, a direct delivery into this folder messed
940 up the contents of the maildirsize file. This was because the regex was
941 used only to exclude .Trash (or whatever) when the size of the mailbox
942 was calculated. There was no check that a delivery was happening into an
943 excluded directory. This bug has been fixed by ignoring all quota
944 processing for deliveries into excluded directories.
946 PH/10 Added the maildirfolder_create_regex option to appendfile.
952 PH/01 The code for finding all the local interface addresses on a FreeBSD
953 system running IPv6 was broken. This may well have applied to all BSD
954 systems, as well as to others that have similar system calls. The broken
955 code found IPv4 interfaces correctly, but gave incorrect values for the
956 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
957 that it would not match correctly against @[] and not recognize the IPv6
960 PH/02 The ipliteral router was not recognizing addresses of the form user@
961 [ipv6:....] because it didn't know about the "ipv6:" prefix.
963 PH/03 Added disable_ipv6.
965 PH/04 Changed $reply_address to use the raw form of the headers instead of the
966 decoded form, because it is most often used to construct To: headers
967 lines in autoreplies, and the decoded form may well be syntactically
968 invalid. However, $reply_address has leading white space removed, and all
969 newlines turned into spaces so that the autoreply transport does not
972 PH/05 If group was specified without a user on a router, and no group or user
973 was specified on a transport, the group from the router was ignored.
975 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
976 for visible compile-time settings that can be used to change these
977 numbers, for those that want even more. Backwards compatibility with old
978 spool files has been maintained. However, going back to a previous Exim
979 release will lost any variables that are in spool files.
981 PH/07 Two small changes when running in the test harness: increase delay when
982 passing a TCP/IP connection to a new process, in case the original
983 process has to generate a bounce, and remove special handling of
984 127.0.0.2 (sic), which is no longer necessary.
986 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
987 be the same on different OS.
989 PH/09 Moved a debug statement in filter processing to avoid a race problem when
992 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
993 whether --show-vars was specified or not
995 JJ/02 exipick: Added support for new ACL variable spool format introduced
998 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
999 syntactically invalid From: or Reply-to: line, and a filter used this to
1000 generate an autoreply, and therefore failed to obtain an address for the
1001 autoreply, Exim could try to deliver to a non-existent relative file
1002 name, causing unrelated and misleading errors. What now happens is that
1003 it logs this as a hard delivery error, but does not attempt to create a
1006 PH/11 The exinext utility has a -C option for testing purposes, but although
1007 the given file was scanned by exinext itself; it wasn't being passed on
1008 when Exim was called.
1010 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1011 an end-of-file indication when reading a command response.
1013 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1014 compiled. In many other places in Exim, IPv6 addresses are always
1015 recognized, so I have changed this. It also means that IPv4 domain
1016 literals of the form [IPV4:n.n.n.n] are now always recognized.
1018 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1019 used if the router is not running as root, for example, when verifying at
1020 ACL time, or when using -bh. The debugging output from this situation was
1021 non-existent - all you got was a failure to exec. I have made two
1024 (a) Failures to set uid/gid, the current directory, or a process leader
1025 in a subprocess such as that created by queryprogram now generate
1026 suitable debugging ouput when -d is set.
1028 (b) The queryprogram router detects when it is not running as root,
1029 outputs suitable debugging information if -d is set, and then runs
1030 the subprocess without attempting to change uid/gid.
1032 PH/15 Minor change to Makefile for building test_host (undocumented testing
1035 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1036 additional section of a DNS packet that returns MX or SRV records.
1037 Instead, it always explicitly searches for A/AAAA records. This avoids
1038 major problems that occur when a DNS server includes only records of one
1039 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1040 fixed another bug: if SRV records were looked up and the corresponding
1041 address records were *not* found in the additional section, the port
1042 values from the SRV records were lost.
1044 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1045 using the correct key (the original address) when searching the retry
1046 rules in order to find which one to use for generating the retry hint.
1048 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1049 adding the default one. Similarly, if it contains a Reply-To: header, the
1050 errors_reply_to option, if set, is not used.
1052 PH/19 When calculating a retry time, Exim used to measure the "time since
1053 failure" by looking at the "first failed" field in the retry record. Now
1054 it does not use this if it is later than than the arrival time of the
1055 message. Instead it uses the arrival time. This makes for better
1056 behaviour in cases where some deliveries succeed, thus re-setting the
1057 "first failed" field. An example is a quota failure for a huge message
1058 when small messages continue to be delivered. Without this change, the
1059 "time since failure" will always be short, possible causing more frequent
1060 delivery attempts for the huge message than are intended.
1061 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1063 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1064 $1, $2, $3) because the numerical variables can be reset during some
1065 expansion items (e.g. "match"), thereby losing the authentication data.
1067 PH/21 Make -bV show the size of off_t variables so that the test suite can
1068 decide whether to run tests for quotas > 2G.
1070 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1071 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1072 filecount value is greater than 2G or if a quota value is greater than 2G
1073 on a system where the size of off_t is not greater than 4, a panic error
1076 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1077 never match. The debug and -bh output now contains an explicit error
1078 message indicating a malformed IPv4 address or mask.
1080 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1081 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1082 PH/23 above applies.
1084 PH/25 Do not write to syslog when running in the test harness. The only
1085 occasion when this arises is a failure to open the main or panic logs
1086 (for which there is an explicit test).
1088 PH/26 Added the /no_tell option to "control=freeze".
1090 PH/27 If a host name lookup failed very early in a connection, for example, if
1091 the IP address matched host_lookup and the reverse lookup yielded a name
1092 that did not have a forward lookup, an error message of the form "no IP
1093 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1094 could be logged. Now it outputs the IP address instead of "NULL".
1096 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1097 allows the sender and the authenticated sender to be set when
1098 submitting a message from within Exim. Since child_open_exim() is
1099 documented for local_scan(), the new function should be too.
1101 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1102 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1103 results in an empty string is now treated as unset.
1105 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1107 PH/31 Added sender_verify_fail logging option.
1109 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1110 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1111 tidied the source and removed it altogether.
1113 PH/33 When a queue run was abandoned because the load average was too high, a
1114 log line was always written; now it is written only if the queue_run log
1115 selector is set. In addition, the log line for abandonment now contains
1116 information about the queue run such as the pid. This is always present
1117 in "start" and "stop" lines but was omitted from the "abandon" line.
1119 PH/34 Omit spaces between a header name and the colon in the error message that
1120 is given when verify = headers_syntax fails (if there are lots of them,
1121 the message gets confusing).
1123 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1124 names, as there are now some PTR records that contain slashes. This check
1125 is only to protect against broken name servers that fall over on strange
1126 characters, so the fact that it applies to all lookups doesn't matter.
1128 PH/36 Now that the new test suite is complete, we can remove some of the
1129 special code in Exim that was needed for the old test suite. For example,
1130 sorting DNS records because real resolvers return them in an arbitrary
1131 order. The new test suite's fake resolver always returns records in the
1134 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1135 bounces) except when queue_only is set, to avoid logging races between
1136 the different processes.
1138 PH/38 Panic-die if .include specifies a non-absolute path.
1140 PH/39 A tweak to the "H" retry rule from its user.
1142 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1143 a label. They prevented compilation on older perls.
1145 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1146 a warning to be raised on newish perls.
1148 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1149 on queue. Changes to match documented behaviour of showing count of
1150 messages matching specified criteria.
1152 PH/40 Changed the default ident timeout from 30s to 5s.
1154 PH/41 Added support for the use of login_cap features, on those BSD systems
1155 that have them, for controlling the resources used by pipe deliveries.
1157 PH/42 The content-scanning code uses fopen() to create files in which to put
1158 message data. Previously it was not paying any attention to the mode of
1159 the files. Exim runs with umask(0) because the rest of the code creates
1160 files with open(), and sets the required mode explicitly. Thus, these
1161 files were ending up world-writeable. This was not a big issue, because,
1162 being within the spool directory, they were not world-accessible. I have
1163 created a function called modefopen, which takes an additional mode
1164 argument. It sets umask(777), creates the file, chmods it to the required
1165 mode, then resets the umask. All the relevant calls to fopen() in the
1166 content scanning code have been changed to use this function.
1168 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1169 to 24 hours. This avoids potential overflow problems when processing G
1170 and H retry rules. I suspect nobody ever tinkers with this value.
1172 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1174 PH/45 When the plaintext authenticator is running as a client, the server's
1175 challenges are checked to ensure they are valid base64 strings. By
1176 default, the authentication attempt is cancelled if an invalid string is
1177 received. Setting client_ignore_invalid_base64 true ignores these errors.
1178 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1179 they are received. Thus, the responses can be made to depend on the
1180 challenges. If an invalid string is ignored, an empty string is placed in
1183 PH/46 Messages that are created by the autoreply transport now contains a
1184 References: header, in accordance with RFCs 2822 and 3834.
1186 PH/47 Added authenticated_sender_force to the smtp transport.
1188 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1190 PH/49 Installed latest patch from the Sieve maintainer.
1192 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1193 was also set, the appendfile transport was unnecessarily scanning a
1194 directory of message files (e.g. for maildir delivery) to find the count
1195 of files (along with the size), even though it did not need this
1196 information. It now does the scan only if it needs to find either the
1197 size of the count of files.
1199 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1201 PH/52 Two bugs concerned with error handling when the smtp transport is
1204 (i) Exim was not creating retry information for temporary errors given
1205 for individual recipients after the DATA command when the smtp transport
1206 was used in LMTP mode. This meant that they could be retried too
1207 frequently, and not timed out correctly.
1209 (ii) Exim was setting the flag that allows error details to be returned
1210 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1211 recipients that were returned after the DATA command.
1213 PH/53 This is related to PH/52, but is more general: for any failing address,
1214 when detailed error information was permitted to be returned to the
1215 sender, but the error was temporary, then after the final timeout, only
1216 "retry timeout exceeded" was returned. Now it returns the full error as
1217 well as "retry timeout exceeded".
1219 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1220 do this, and (what is worse) MTAs that accept it.
1222 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1223 will now be deprecated.
1225 PH/56 New os.c-cygwin from the Cygwin maintainer.
1227 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1228 formats (previously only available in exim formats via -bpr, -bpru,
1229 and -bpra. Now also available in native and exiqgrep formats)
1231 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1232 with very large, slow to parse queues
1234 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1236 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1238 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1239 responses to authentication challenges, though it was showing the
1240 challenges; (ii) I've removed the CR characters from the debug output for
1243 PH/58 Allow for the insertion of a newline as well as a space when a string
1244 is turned into more than one encoded-word during RFC 2047 encoding. The
1245 Sieve code now uses this.
1247 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1248 data_4xx, lost_connection, tls_required.
1250 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1251 message was being sent as an SMTP response.
1253 PH/61 Add -l and -k options to exicyclog.
1255 PH/62 When verifying, if an address was redirected to one new address, so that
1256 verification continued, and the new address failed or deferred after
1257 having set something in $address_data, the value of $address_data was not
1258 passed back to the ACL. This was different to the case when no
1259 redirection occurred. The value is now passed back in both cases.
1261 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1262 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1263 use login_cap.h, so on its own it isn't the distinguishing feature. The
1264 new name refers directly to the setclassresources() function.
1266 PH/65 Added configuration files for NetBSD3.
1268 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1270 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1271 is preferred over IPv4.
1273 PH/68 The bounce_return_message and bounce_return_body options were not being
1274 honoured for bounces generated during the reception of non-SMTP messages.
1275 In particular, this applied to messages rejected by the ACL. This bug has
1276 been fixed. However, if bounce_return_message is true and bounce_return_
1277 body is false, the headers that are returned for a non-SMTP message
1278 include only those that have been read before the error was detected.
1279 (In the case of an ACL rejection, they have all been read.)
1281 PH/69 The HTML version of the specification is now built in a directory called
1282 spec_html instead of spec.html, because the latter looks like a path with
1283 a MIME-type, and this confuses some software.
1285 PH/70 Catch two compiler warnings in sieve.c.
1287 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1288 function verify_get_ident() calls ip_connect() to connect a socket, but
1289 if the "connect()" function timed out, ip_connect() used to close the
1290 socket. However, verify_get_ident() also closes the socket later, and in
1291 between Exim writes to the log, which may get opened at this point. When
1292 the socket was closed in ip_connect(), the log could get the same file
1293 descriptor number as the socket. This naturally causes chaos. The fix is
1294 not to close the socket in ip_connect(); the socket should be closed by
1295 the function that creates it. There was only one place in the code where
1296 this was missing, in the iplookup router, which I don't think anybody now
1297 uses, but I've fixed it anyway.
1299 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1300 well as to direct DNS lookups. Otherwise the handling of names in host
1301 lists is inconsistent and therefore confusing.
1307 PH/01 Two changes to the default runtime configuration:
1309 (1) Move the checks for relay_from_hosts and authenticated clients from
1310 after to before the (commented out) DNS black list checks.
1312 (2) Add control=submission to the relay_from_hosts and authenticated
1313 clients checks, on the grounds that messages accepted by these
1314 statements are most likely to be submissions.
1316 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1318 (1) Generate an error if the third argument for the ${prvs expansion is
1321 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1324 (3) Reset the variables that are obtained from the first argument of
1325 ${prvscheck and used in the second argument before leaving the code,
1326 because their memory is reclaimed, so using them afterwards may do
1329 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1330 one (it's much easier than Tom thought :-).
1332 (5) Because of (4), we can now allow for the use of $prvscheck_result
1333 inside the third argument.
1335 PH/03 For some reason, the default setting of PATH when running a command from
1336 a pipe transport was just "/usr/bin". I have changed it to
1339 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1340 anything to be listed in the output from -bV.
1342 PH/05 When a filter generated an autoreply, the entire To: header line was
1343 quoted in the delivery log line, like this:
1345 => >A.N.Other <ano@some.domain> <original@ddress> ...
1347 This has been changed so that it extracts the operative address. There
1348 may be more than one such address. If so, they are comma-separated, like
1351 => >ano@some.domain,ona@other.domain <original@ddress> ...
1353 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1354 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1355 not being looked up in the rDNS to get a host name, Exim was showing the
1356 IP address twice in Received: lines, even though the IP addresses were
1357 identical. For example:
1359 Received: from [1.2.3.4] (helo=[1.2.3.4])
1361 However, if the real host name was known, it was omitting the HELO data
1362 if it matched the actual IP address. This has been tidied up so that it
1363 doesn't show the same IP address twice.
1365 PH/07 When both +timestamp and +memory debugging was on, the value given by
1366 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1367 called by the debug printing, thereby overwriting the timestamp buffer.
1368 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1370 PH/08 When the original message was included in an autoreply transport, it
1371 always said "this is a copy of the message, including all the headers",
1372 even if body_only or headers_only was set. It now gives an appropriate
1375 PH/09 Applied a patch from the Sieve maintainer which:
1377 o fixes some comments
1378 o adds the (disabled) notify extension core
1379 o adds some debug output for the result of if/elsif tests
1380 o points to the current vacation draft in the documentation
1381 and documents the missing references header update
1385 o fixes a bug in processing the envelope test (when testing
1386 multiple envelope elements, the last element determinted the
1389 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1390 Electronic Mail") by including:
1392 Auto-submitted: auto-generated
1394 in the messages that it generates (bounce messages and others, such as
1395 warnings). In the case of bounce messages for non-SMTP mesages, there was
1396 also a typo: it was using "Auto_submitted" (underscore instead of
1397 hyphen). Since every message generated by Exim is necessarily in response
1398 to another message, thes have all been changed to:
1400 Auto-Submitted: auto-replied
1402 in accordance with these statements in the RFC:
1404 The auto-replied keyword:
1406 - SHOULD be used on messages sent in direct response to another
1407 message by an automatic process,
1409 - MUST NOT be used on manually-generated messages,
1411 - MAY be used on Delivery Status Notifications (DSNs) and Message
1412 Disposition Notifications (MDNs),
1414 - MUST NOT be used on messages generated by automatic or periodic
1415 processes, except for messages which are automatic responses to
1418 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1419 to the default Received: header definition.
1421 PH/12 Added log selector acl_warn_skipped (default on).
1423 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1424 variables because (a) they are in the wrong storage pool and (b) even if
1425 they were copied, it wouldn't work properly because of the caching.
1427 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1428 checking when decoding. Apparently there are clients that generate
1429 overlong encoded strings. Why am I not surprised?
1431 PH/15 If the first argument of "${if match_address" was not empty, but did not
1432 contain an "@" character, Exim crashed. Now it writes a panic log message
1433 and treats the condition as false.
1435 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1437 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1438 extension "envelope-auth". The code is finished and in agreement with
1439 other implementations, but there is no documentation so far and in fact,
1440 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1441 not changing the active code.
1443 Print executed "if" and "elsif" statements when debugging is used. This
1444 helps a great deal to understand what a filter does.
1446 Document more things not specified clearly in RFC3028. I had all this
1447 sorted out, when out of a sudden new issues came to my mind. Oops."
1449 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1452 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1453 canonical form (as documented). However, after a host name lookup from
1454 the IP address, check_host() was doing a simple string comparison with
1455 addresses acquired from the DNS when checking that the found name did
1456 have the original IP as one of its addresses. Since any found IPv6
1457 addresses are likely to be in abbreviated form, the comparison could
1458 fail. Luckily, there already exists a function for doing the comparison
1459 by converting both addresses to binary, so now that is used instead of
1460 the text comparison.
1462 PH/20 There was another similar case to PH/19, when a complete host name was
1463 given in a host list; looking up its IP address could give an abbreviated
1464 form, whereas the current host's name might or might not be abbreviated.
1465 The same fix has been applied.
1471 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1472 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1475 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1476 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1478 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1480 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1481 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1482 Also, for "Auto-Submitted", treat anything other than "no" as
1483 non-personal, in accordance with RFC 3834. (Previously it treated
1484 anything starting "auto-" as non-personal.)
1486 TF/01 The control=submission/name=... option had a problem with syntax
1487 errors if the name included a slash character. The /name= option
1488 now slurps the rest of the string, so it can include any characters
1489 but it must come last in the list of options (after /sender_retain
1492 PH/05 Some modifications to the interface to the fake nameserver for the new
1500 TK/01 Added the "success_on_redirect" address verification option. See
1501 NewStuff for rationale and an example.
1503 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1505 PH/02 Patch to exigrep to allow it to work on syslog lines.
1507 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1508 fread() to skip over the body file's header line, because in Cygwin the
1509 header line is locked and is inaccessible.
1511 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1512 co-exist for some time) to make it clear that it is the Exim ID that is
1513 referenced, not the Message-ID: header line.
1515 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1516 string_format() function, because snprintf() does not exist on all
1519 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1520 ${stat: expansion item.
1522 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1523 protocol synchronization error", to keep the pedants happy.
1525 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1526 well as for IRIX systems, when gcc is being used. See the host.c source
1529 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1531 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1534 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1535 transport and to the smtp transport in LMTP mode.
1537 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1539 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1540 run a filter in a subprocess. This could lead to confusion in subsequent
1541 lookups in the parent process. There should also be a search_tidyup() at
1542 the end of the subprocess.
1544 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1545 only if the host matched helo_try_verify_hosts, which caused the
1546 verification to occur when the EHLO/HELO command was issued. The ACL just
1547 tested the remembered result. Now, if a previous verification attempt has
1548 not happened, "verify = helo" does it there and then.
1550 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1552 TK/03 Fix log output including CR from clamd.
1554 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1555 exist provoked a memory error which could cause a segfault.
1557 PH/15 Installed PCRE 6.2
1559 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1561 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1562 of the problem. Specifically, suggested +O2 rather than +O1 for the
1565 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1567 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1568 record, the port was not correctly logged when the outgoing_port log
1569 selector was set (it logged the transort's default port).
1571 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1572 fallback_hosts, and "hosts" in the smtp transport.
1574 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1575 host errors such as "Connection refused".
1577 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1578 authentication with radiusclient 0.4.9:
1580 - Error returned from rc_read_config was caught wrongly
1581 - Username/password not passed on to radius server due to wrong length.
1583 The presumption is that some radiusclient API changes for 4.51/PH/17
1584 were not taken care of correctly. The code is still untested by me (my
1585 Linux distribution still has 0.3.2 of radiusclient), but it was
1586 contributed by a Radius user.
1588 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1589 expanding the "port" option of the smtp transport.
1591 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1592 while reading a MIME header. Thanks to Tom Hughes for a patch.
1594 PH/24 Include config.h inside local_scan.h so that configuration settings are
1597 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1598 that in an ACL for RCPT (for example), you can examine exactly what was
1601 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1602 commands, but it was not correctly comparing the address with the actual
1603 client host address. Thus, it would show the EHLO address in Received:
1604 header lines when this was not necessary.
1606 PH/27 Added the % operator to ${eval:}.
1608 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1609 it should be ignoring failures (because with -C, for example, it has lost
1610 privilege). It wasn't ignoring creation failures other than "already
1613 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1616 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1617 because an input line was too long, either on its own, or by virtue of
1618 too many continuations, the temporary file was not being removed, and the
1619 return code was incorrect.
1621 PH/31 Missing "BOOL" in function definition in filtertest.c.
1623 PH/32 Applied Sieve patches from the maintainer.
1625 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1627 PH/33 Added "verify = not_blind".
1629 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1630 Local/Makefile (with some defaults set). These are used in built scripts
1631 such as exicyclog, but they have never been used in the exim_install
1632 script (though there are many overriding facilities there). I have
1633 arranged that the exim_install script now takes note of these two
1636 PH/35 Installed configuration files for Dragonfly.
1638 PH/36 When a locally submitted message by a trusted user did not contain a
1639 From: header, and the sender address was obtained from -f or from an SMTP
1640 MAIL command, and the trusted user did not use -F to supply a sender
1641 name, $originator_name was incorrectly used when constructing a From:
1642 header. Furthermore, $originator_name was used for submission mode
1643 messages from external hosts without From: headers in a similar way,
1644 which is clearly wrong.
1646 PH/37 Added control=suppress_local_fixups.
1648 PH/38 When log_selector = +received_sender was set, and the addition of the
1649 sender made the log line's construction buffer exactly full, or one byte
1650 less than full, an overflow happened when the terminating "\n" was
1653 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1654 when the result of a list match is failure because a DNS lookup failed.
1656 PH/40 RM_COMMAND is now used in the building process.
1658 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1659 the "build-* directories that it finds.
1661 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1662 address in a domain literal was a prefix of an interface address.
1664 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1665 when verifying a sender address, unless rewrite_headers is false.
1667 PH/44 Wrote a long comment about why errors_to addresses are verified as
1668 recipients, not senders.
1670 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1671 the ratelimit ACL was added.
1673 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1675 PH/46 Added extra information about PostgreSQL errors to the error string.
1677 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1678 avoiding the need to install special zones in a real server. This is
1679 backwards compatible; if it can't find the fake resolver, it drops back.
1680 Thus, both old and new test suites can be run.
1682 TF/02 Added util/ratelimit.pl
1684 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1685 clock is set back in time.
1687 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1688 Candler <B.Candler@pobox.com>.
1690 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1691 OK for addresses that are the result of redirections.
1693 PH/48 A number of further additions for the benefit of the new test suite,
1694 including a fake gethostbyname() that interfaces to the fake DNS resolver
1697 TF/06 The fix for widen_domains has also been applied to qualify_single and
1698 search_parents which are the other dnslookup options that can cause
1701 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1704 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1706 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1707 service" if it encountered an MX record with an empty target hostname.
1708 The message is now "an MX or SRV record indicated no SMTP service".
1710 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1711 if the DNS of the sending site is misconfigured. This is quite a
1712 common situation. This change restores the behaviour of treating a
1713 helo verification defer as a failure.
1715 PH/51 If self=fail was set on a router, the bounce message did not include the
1716 actual error message.
1722 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1724 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1725 command itself ended in error, the underlying message about the transport
1726 filter timeout was being overwritten with the pipe command error. Now the
1727 underlying error message should be appended to the second error message.
1729 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1731 PH/02 Reduce the amount of output that "make" produces by default. Full output
1732 can still be requested.
1734 PH/03 The warning log line about a condition test deferring for a "warn" verb
1735 was being output only once per connection, rather than after each
1736 occurrence (because it was using the same function as for successful
1737 "warn" verbs). This seems wrong, so I have changed it.
1739 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1740 it should not have, which might have caused a crash in the right
1741 circumstances, but probably never did.
1743 PH/04 Installed a modified version of Tony Finch's patch to make submission
1744 mode fix the return path as well as the Sender: header line, and to
1745 add a /name= option so that you can make the user's friendly name appear
1748 TF/03 Added the control = fakedefer ACL modifier.
1750 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1751 Mark Lowes for thorough testing.
1753 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1755 TK/03 Merged latest SRS patch from Miles Wilton.
1757 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1758 with the definition in sysexits.h (which is #included earlier).
1759 Fortunately, Exim does not actually use EX_OK. The code used to try to
1760 preserve the sysexits.h value, by assumimg that macro definitions were
1761 scanned for macro replacements. I have been disabused of this notion,
1762 so now the code just undefines EX_OK before #including unistd.h.
1764 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1765 in the smtp transport. When a block could not be written in a single
1766 write() function, the timeout was being re-applied to each part-write.
1767 This seems wrong - if the receiver was accepting one byte at a time it
1768 would take for ever. The timeout is now adjusted when this happens. It
1769 doesn't have to be particularly precise.
1771 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1772 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1774 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1775 without a domain if the check to <postmaster@domain> fails.
1777 SC/01 Eximstats: added -xls and the ability to specify output files
1778 (patch written by Frank Heydlauf).
1780 SC/02 Eximstats: use FileHandles for outputing results.
1782 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1784 SC/04 Eximstats: fixed display of large numbers with -nvr option
1786 SC/05 Eximstats: fixed merging of reports with empty tables.
1788 SC/06 Eximstats: added the -include_original_destination flag
1790 SC/07 Eximstats: removed tabs and trailing whitespace.
1792 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1794 TK/06 MBOX spool code: Add real "From " MBOX separator line
1795 so the .eml file is really in mbox format (even though
1796 most programs do not really care). Patch from Alex Miller.
1798 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1799 The latter is generated from $received_to and is only set if the
1800 message has one envelope recipient. SA can use these headers,
1801 obviously out-of-the-box. Patch from Alex Miller.
1803 PH/08 The ${def test on a variable was returning false if the variable's
1804 value was "0", contrary to what the specification has always said!
1805 The result should be true unless the variable is empty.
1807 PH/09 The syntax error of a character other than { following "${if
1808 def:variable_name" (after optional whitespace) was not being diagnosed.
1809 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1810 accidental colon was present, for example, could give incorrect results.
1812 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1813 result is used (not including appendfile, where other changes are about
1816 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1817 This involved changing a lot of size variables from int to off_t. It
1818 should work with maildirs and everything.
1820 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1821 spamd dying while we are connected to it.
1823 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1826 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1827 function for BDB 4.3.
1829 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1831 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1832 the sources for PCRE have been re-arranged and more files are now
1835 PH/15 The code I had for printing potentially long long variables in PH/11
1836 above was not the best (it lost precision). The length of off_t variables
1837 is now inspected at build time, and an appropriate printing format (%ld
1838 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1839 to be "long long int" or "long int". This is needed for the internal
1840 formatting function string_vformat().
1842 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1843 the configuration file to be ":syslog", then the script "guesses" where
1844 the logs files are, rather than using the compiled in default. In our
1845 case the guess is not the same as the compiled default, so the script
1846 suddenly stopped working when I started to use syslog. The patch checks
1847 to see if log_file_path is "". If so, it attempts to read it from exim
1848 with no configuration file to get the compiled in version, before it
1849 falls back to the previous guessing code."
1851 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1852 implementing BATV in an Exim configuration. See NewStuff for the gory
1855 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1856 Makefile that are specific to HP-UX.
1858 PH/18 If the "use_postmaster" option was set for a recipient callout together
1859 with the "random" option, the postmaster address was used as the MAIL
1860 FROM address for the random test, but not for the subsequent recipient
1861 test. It is now used for both.
1863 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1864 patch removes a few documentation additions to RFC 3028, because the
1865 latest draft now contains them. It adds the new en;ascii-case comparator
1866 and a new error check for 8bit text in MIME parts. Comparator and
1867 require names are now matched exactly. I enabled the subaddress
1868 extension, but it is not well tested yet (read: it works for me)."
1870 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1871 rework some of the code of TK/09 above to avoid the hardwired use of
1872 "%lld" and "long long". Replaced the call to snprintf() with a call to
1875 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1876 records point to non-existent hosts", "retry timeout exceeded", and
1877 "retry time not reached for any host after a long failure period".
1879 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1880 experimental DomainKeys support:
1882 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1883 (2) On an error such as an illegally used "control", the wrong name for
1884 the control was given.
1886 These problems did NOT occur unless DomainKeys support was compiled.
1888 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1890 PH/24 Added ${if match_ip condition.
1892 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1893 databases so that it will be absolutely obvious if a crash occurs in the
1894 DB library. This is a regular occurrence (often caused by mis-matched
1897 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1898 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1899 picked up on a user's system that detects such things. There doesn't seem
1900 to be a gcc warning option for this - only an attribute that has to be
1901 put on the function's prototype. It seems that in Fedora Core 4 they have
1902 set this on a number of new functions. No doubt there will be more in due
1905 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1906 specify a transport. However, if an address that was verified by such a
1907 router was the subject of a callout, Exim crashed because it tried to
1908 read the rcpt_include_affixes from the non-existent transport. Now it
1909 just assumes that the setting of that option is false. This bug was
1910 introduced by 4.51/PH/31.
1912 PH/28 Changed -d+all to exclude +memory, because that information is very
1913 rarely of interest, but it makes the output a lot bigger. People tend to
1914 do -d+all out of habit.
1916 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1917 code in os-type was giving problems when libc.so lives in lib64, like on
1920 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1921 aren't the modern standard, and it seems that some systems' include files
1922 don't always have them. Exim was already checking for some of the newer
1923 ones like T_AAAA, and defining it itself. I've added checks for all the
1924 record types that Exim uses.
1926 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1927 not automatically generating a new one, as it is supposed to. This
1928 prevented TLS from working. If the file did exist, but contained invalid
1929 data, a new version was generated, as expected. It was only the case of a
1930 non-existent file that was broken.
1932 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1933 with a change in libdomainkeys > 0.64.
1935 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1936 from DNS. If the selector record carries the flag, it now has
1937 precedence over the domain-wide flag.
1939 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1941 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1942 the use of a port name that isn't defined in /etc/services) occurred, the
1943 message was deferred as in a normal delivery, and thus remained on the
1944 spool, instead of being failed because of the mua_wrapper setting. This
1945 is now fixed, and I tidied up some of the mua_wrapper messages at the
1948 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1949 the messages in a hash of arrays rather than using individual hashes.
1950 This is a bit cleaner and results in dramatic memory savings, albeit
1951 at a slight CPU cost.
1953 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1954 as requested by Marc Sherman.
1956 SC/10 Eximstats: added histograms for user specified patterns as requested
1959 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1961 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1962 fopen() in the content-scanning modules that did not already have it.
1968 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1969 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1971 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1973 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1975 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1976 to test Sieve filters that use "vacation".
1978 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1979 that changes the way the GnuTLS parameters are stored in the cache file.
1980 The new format can be generated externally. For backward compatibility,
1981 if the data in the cache doesn't make sense, Exim assumes it has read an
1982 old-format file, and it generates new data and writes a new file. This
1983 means that you can't go back to an older release without removing the
1986 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1987 work if there are any delivery delays because "one_time" forces the
1988 parent to be marked "delivered", so its unseen clone is never tried
1989 again. For this reason, Exim now forbids the simultaneous setting of
1992 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1993 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1994 entry if you want to know the details. The fix, however, neglected to
1995 consider the case when local delivery batching is involved. The test for
1996 "previously delivered" was not happening when checking to see if an
1997 address could be batched with a previous (undelivered) one; under
1998 certain circumstances this could lead to multiple deliveries to the same
2001 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2002 in its include files, and this causes problems building Exim.
2004 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2005 header_syntax) but Exim was just ignoring anything given after a slash.
2006 In particular, this caused confusion with an attempt to use "verify =
2007 reverse_host_lookup/defer_ok". An error is now given when options are
2008 supplied for verify items that do not have them. (Maybe reverse_host_
2009 lookup should have a defer_ok option, but that's a different point.)
2011 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2012 defined by RFC 821) to 2048, because there were problems with some AUTH
2013 commands, and RFC 1869 says the size should be increased for extended
2014 SMTP commands that take arguments.
2016 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2019 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2020 "unknown" error; now it says that the functionality isn't in the binary.
2022 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2023 an address' error message when a string expansion fails (syntax or
2024 whatever). Otherwise the password may appear in the log. Following change
2025 PH/42 below, there is no longer a chance of it appearing in a bounce
2028 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2030 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2031 of its addresses was ever tried. (Bugzilla bug #2.)
2033 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2034 the result incorrectly in the debug output. (It correctly added a newline
2035 to what was transported.)
2037 TF/01 Added $received_time.
2039 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2040 commented out examples of how to interface to a virus scanner and to
2041 SpamAssassin. Also added commented examples of av_scanner and
2042 spamd_address settings.
2044 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2045 and controls are allowed in which ACLs. There were a couple of minor
2046 errors. Some of the entries in the conditions table (which is a table of
2047 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2048 negation of where the condition IS allowed.
2050 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2052 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2053 header file does not have a version number, so I've had to invent a new
2054 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2055 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2056 radiusclient), but it was contributed by a Radius user.
2058 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2059 files or directories, for OpenSSL.
2061 PH/19 When an Exim process that is running as root has to create an Exim log
2062 file, it does so in a subprocess that runs as exim:exim so as to get the
2063 ownership right at creation (otherwise, other Exim processes might see
2064 the file with the wrong ownership). There was no test for failure of this
2065 fork() call, which would lead to the process getting stuck as it waited
2066 for a non-existent subprocess. Forks do occasionally fail when resources
2067 run out. I reviewed all the other calls to fork(); they all seem to check
2070 PH/20 When checking for unexpected SMTP input at connect time (before writing
2071 the banner), Exim was not dealing correctly with a non-positive return
2072 from the read() function. If the client had disconnected by this time,
2073 the result was a log entry for a synchronization error with an empty
2074 string after "input=" when read() returned zero. If read() returned -1
2075 (an event I could not check), uninitialized data bytes were printed.
2076 There were reports of junk text (parts of files, etc) appearing after
2079 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2081 PH/22 Added support for macro redefinition, and (re)definition in between
2082 driver and ACL definitions.
2084 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2085 forgetting to use the resulting value; it was using the unexpanded value.
2087 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2088 hadn't been configured. The fix is from Juergen Kreileder, who
2089 understands it better than I do:
2091 "Here's what I see happening with three configured cyrus_sasl
2092 authenticators configured (plain, login, cram-md5):
2094 On startup auth_cyrus_sasl_init() gets called for each of these.
2095 This means three calls to sasl_listmech() without a specified mech_list.
2096 => SASL tests which mechs of all available mechs actually work
2097 => three warnings about OTP not working
2098 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2100 With the patch, sasl_listmech() also gets called three times. But now
2101 SASL's mech_list option is set to the server_mech specified in the the
2102 authenticator. Or in other words, the answer from sasl_listmech()
2103 gets limited to just the mech you're testing for (which is different
2105 => the return list contains just 'plain' or 'login', 'cram-md5' or
2106 nothing depending on the value of ob->server_mech.
2108 I've just tested the patch: Authentication still works fine,
2109 unavailable mechs specified in the exim configuration are still
2110 caught, and the auth.log warnings about OTP are gone."
2112 PH/25 When debugging is enabled, the contents of the command line are added
2113 to the debugging output, even when log_selector=+arguments is not
2116 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2117 answer is "GNU", and only if the return is "GNU/something" is the answer
2120 PH/27 $acl_verify_message is now set immediately after the failure of a
2121 verification in an ACL, and so is available in subsequent modifiers. In
2122 particular, the message can be preserved by coding like this:
2124 warn !verify = sender
2125 set acl_m0 = $acl_verify_message
2127 Previously, $acl_verify_message was set only while expanding "message"
2128 and "log_message" when a very denied access.
2130 PH/28 Modified OS/os.c-Linux with
2132 -#ifndef OS_LOAD_AVERAGE
2133 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2135 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2136 nomenclature these days.)
2138 PH/29 Installed patch from the Sieve maintainer that adds the options
2139 sieve_useraddress and sieve_subaddress to the redirect router.
2141 PH/30 In these circumstances:
2142 . Two addresses routed to the same list of hosts;
2143 . First host does not offer TLS;
2144 . First host accepts first address;
2145 . First host gives temporary error to second address;
2146 . Second host offers TLS and a TLS session is established;
2147 . Second host accepts second address.
2148 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2149 and peerdn, if requested) that were in fact used only for the second
2152 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2153 attention to any local part prefix or suffix that was matched by the
2154 router that accepted the address. It now behaves in the same way as it
2155 does for delivery: the affixes are removed from the local part unless
2156 rcpt_include_affixes is set on the transport.
2158 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2159 timeout during the DATA phase of an incoming message.
2161 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2162 applied a patch sanctioned by the Sieve maintainer.
2164 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2165 the uid or gid is negative. A case of a negative gid caused this to be
2166 noticed. The fix allows for either to be negative.
2168 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2169 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2172 PH/36 The change PH/12 above was broken. Fixed it.
2174 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2175 the grounds that routing the same address twice would always produce the
2176 same answer. This might have been true once, but it is certainly no
2177 longer true now. Routing a child address may depend on the previous
2178 routing that produced that child. Some complicated redirection strategies
2179 went wrong when messages had multiple recipients, and made Exim's
2180 behaviour dependent on the order in which the addresses were given.
2182 I have moved the duplicate checking until after the routing is complete.
2183 Exim scans the addresses that are assigned to local and remote
2184 transports, and removes any duplicates. This means that more work will be
2185 done, as duplicates will always all be routed, but duplicates are
2186 presumably rare, so I don't expect this is of any significance.
2188 For deliveries to pipes, files, and autoreplies, the duplicate checking
2189 still happens during the routing process, since they are not going to be
2192 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2193 It corrects a timeout issue with spamd. This is Ian's comment: "The
2194 background is that sometimes spamd either never reads data from a
2195 connection it has accepted, or it never writes response data. The exiscan
2196 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2197 blindly assumes that writes won't block so it may never time out."
2199 PH/39 Allow G after quota size as well as K and M.
2201 PH/40 The value set for $authenticated_id in an authenticator may not contain
2202 binary zeroes or newlines because the value is written to log lines and
2203 to spool files. There was no check on this. Now the value is run through
2204 the string_printing() function so that such characters are converted to
2205 printable escape sequences.
2207 PH/41 $message_linecount is a new variable that contains the total number of
2208 lines in the message. Compare $body_linecount, which is the count for the
2211 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2212 bounce and delay warning messages, except in certain special cases, which
2215 (a) An SMTP error message from a remote host;
2216 (b) A message specified in a :fail: redirection;
2217 (c) A message specified in a "fail" command in a system filter;
2218 (d) A message specified in a FAIL return from the queryprogram router;
2219 (e) A message specified by the cannot_route_message router option.
2221 In these cases only, Exim does include the error details in bounce and
2222 warning messages. There are also a few cases where bland messages such
2223 as "unrouteable address" or "local delivery error" are given.
2225 PH/43 $value is now also set for the "else" part of a ${run expansion.
2227 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2228 being worked on, but at least Exim now implements the latest version to
2231 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2232 process to complete was treated as a delivery failure, a timeout while
2233 writing the message to the pipe was logged, but erroneously treated as a
2234 successful delivery. Such timeouts include transport filter timeouts. For
2235 consistency with the overall process timeout, these timeouts are now
2236 treated as errors, giving rise to delivery failures by default. However,
2237 there is now a new Boolean option for the pipe transport called
2238 timeout_defer, which, if set TRUE, converts the failures into defers for
2239 both kinds of timeout. A transport filter timeout is now identified in
2242 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2243 systems where "make" and "gmake" are different, calling "gmake" at top
2244 level broke things. I've arranged for the value of $(MAKE) to be passed
2245 from the Makefile to this script so that it can call the same version of
2249 A note about Exim versions 4.44 and 4.50
2250 ----------------------------------------
2252 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2253 changes of various kinds. As a consequence, a big documentation update was
2254 needed. This delayed the release for rather longer than seemed good, especially
2255 in the light of a couple of (minor) security issues. Therefore, the changes
2256 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2257 release. So 4.44 and 4.50 are in effect two different branches that both start
2260 I have left the 4.50 change log unchanged; it contains all the changes since
2261 4.43. The change log for 4.44 is below; many of its items are identical to
2262 those for 4.50. This seems to be the most sensible way to preserve the
2263 historical information.
2269 1. Minor wording change to the doc/README.SIEVE file.
2271 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2272 computation of the current number of files was incorrect.
2274 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2275 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2278 4. Give more explanation in the error message when the command for a transport
2279 filter fails to execute.
2281 5. There are several places where Exim runs a non-Exim command in a
2282 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2283 was being done only for the command run by the queryprogram router. It is
2284 now done for all such subprocesses. The other cases are: ${run, transport
2285 filters, and the commands run by the lmtp and pipe transports.
2287 6. Added CONFIGURE_GROUP build-time option.
2289 7. Some older OS have a limit of 256 on the maximum number of file
2290 descriptors. Exim was using setrlimit() to set 1000 as a large value
2291 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2292 systems. I've change it so that if it can't get 1000, it tries for 256.
2294 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2295 was an oversight, and furthermore, ever since the addition of extra
2296 controls (e.g. 4.43/32), the checks on when to allow different forms of
2297 "control" were broken. There should now be diagnostics for all cases when a
2298 control that does not make sense is encountered.
2300 9. Added the /retain_sender option to "control=submission".
2302 10. $recipients is now available in the predata ACL (oversight).
2304 11. Tidy the search cache before the fork to do a delivery from a message
2305 received from the command line. Otherwise the child will trigger a lookup
2306 failure and thereby defer the delivery if it tries to use (for example) a
2307 cached ldap connection that the parent has called unbind on.
2309 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2310 of $address_data from the recipient verification was clobbered by the
2311 sender verification.
2313 13. The value of address_data from a sender verification is now available in
2314 $sender_address_data in subsequent conditions in the ACL statement.
2316 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2318 15. Added a new option "connect=<time>" to callout options, to set a different
2321 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2322 was its contents. (It was OK if the option was not defined at all.)
2324 17. A "Completed" log line is now written for messages that are removed from
2325 the spool by the -Mrm option.
2327 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2328 information about exactly what failed.
2330 19. Added -dd to debug only the daemon process.
2332 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2333 handles timeouts, both on the server side and network timeouts. Renamed the
2334 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2336 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2337 It is now set to "smtps".
2339 22. $host_address is now set to the target address during the checking of
2340 ignore_target_hosts.
2342 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2343 being passed; this would have caused $sender_host_name to have been used if
2344 matching the list had actually called for a host name (not very likely,
2345 since this list is usually IP addresses). A host name is now passed as
2348 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2349 code that specifies a non-restarting handler (typically sigaction() in
2350 modern systems) in an attempt to fix a rare and obscure crash bug.
2352 25. Narrowed the window for a race in the daemon that could cause it to ignore
2353 SIGCHLD signals. This is not a major problem, because they are used only to
2354 wake it up if nothing else does.
2356 26. A malformed maildirsize file could cause Exim to calculate negative values
2357 for the mailbox size or file count. Odd effects could occur as a result.
2358 The maildirsize information is now recalculated if the size or filecount
2361 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2362 support for a long time. Removed HAVE_SYS_VFS_H.
2364 28. Installed the latest version of exipick from John Jetmore.
2366 29. In an address list, if the pattern was not a regular expression, an empty
2367 subject address (from a bounce message) matched only if the pattern was an
2368 empty string. Non-empty patterns were not even tested. This was the wrong
2369 because it is perfectly reasonable to use an empty address as part of a
2370 database query. An empty address is now tested by patterns that are
2371 lookups. However, all the other forms of pattern expect the subject to
2372 contain a local part and a domain, and therefore, for them, an empty
2373 address still always fails if the pattern is not itself empty.
2375 30. Exim went into a mad DNS loop when attempting to do a callout where the
2376 host was specified on an smtp transport, and looking it up yielded more
2377 than one IP address.
2379 31. Re-factored the code for checking spool and log partition space into a
2380 function that finds that data and another that does the check. The former
2381 is then used to implement four new variables: $spool_space, $log_space,
2382 $spool_inodes, and $log_inodes.
2384 32. The RFC2047 encoding function was originally intended for short strings
2385 such as real names; it was not keeping to the 75-character limit for
2386 encoded words that the RFC imposes. It now respects the limit, and
2387 generates multiple encoded words if necessary. To be on the safe side, I
2388 have increased the buffer size for the ${rfc2047: expansion operator from
2391 33. It is now permitted to omit both strings after an "if" condition; if the
2392 condition is true, the result is "true". As before, when the second string
2393 is omitted, a false condition yields an empty string. This makes it less
2394 cumbersome to write custom ACL and router conditions.
2396 34. Failure to deliver a bounce message always caused it to be frozen, even if
2397 there was an errors_to setting on the router. The errors_to setting is now
2400 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2401 canonical form (fully expanded) before being placed in
2402 $sender_host_address.
2404 36. The table in the code that translates DNS record types into text (T_A to
2405 "A" for instance) was missing entries for NS and CNAME. It is just possible
2406 that this could have caused confusion if both these types were looked up
2407 for the same domain, because the text type is used as part of Exim's
2408 per-process caching. But the chance of anyone hitting this buglet seems
2411 37. The dnsdb lookup has been extended in a number of ways.
2413 (1) There is a new type, "zns", which walks up the domain tree until it
2414 finds some nameserver records. It should be used with care.
2416 (2) There is a new type, "mxh", which is like "mx" except that it returns
2417 just the host names, not the priorities.
2419 (3) It is now possible to give a list of domains (or IP addresses) to be
2420 looked up. The behaviour when one of the lookups defers can be
2421 controlled by a keyword.
2423 (4) It is now possible to specify the separator character for use when
2424 multiple records are returned.
2426 38. The dnslists ACL condition has been extended: it is now possible to supply
2427 a list of IP addresses and/or domains to be looked up in a particular DNS
2430 39. Added log_selector=+queue_time_overall.
2432 40. When running the queue in the test harness, wait just a tad after forking a
2433 delivery process, to get repeatability of debugging output.
2435 41. Include certificate and key file names in error message when GnuTLS fails
2436 to set them up, because the GnuTLS error message doesn't include the name
2437 of the failing file when there is a problem reading it.
2439 42. Allow both -bf and -bF in the same test run.
2441 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2443 44. The "Exiscan patch" is now merged into the mainline Exim source.
2445 45. Sometimes the final signoff response after QUIT could fail to get
2446 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2447 < 0 before doing a fflush(). This bug looks as though it goes back to the
2448 introduction of TLS in release 3.20, but "sometimes" must have been rare
2449 because the tests only now provoked it.
2451 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2452 (this can affect the format of dates).
2454 47. exim_tidydb, when checking for the continued existence of a message for
2455 which it has found a message-specific retry record, was not finding
2456 messages that were in split spool directories. Consequently, it was
2457 deleting retry records that should have stayed in existence.
2459 48. Steve fixed some bugs in eximstats.
2461 49. The SPA authentication driver was not abandoning authentication and moving
2462 on to the next authenticator when an expansion was forced to fail,
2463 contradicting the general specification for all authenticators. Instead it
2464 was generating a temporary error. It now behaves as specified.
2466 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2467 (the order specifies the preference for clients). The order is now AES256,
2468 AES128, 3DES, ARCFOUR128.
2470 51. Small patch to Sieve code - explicitly set From: when generating an
2473 52. Exim crashed if a remote delivery caused a very long error message to be
2474 recorded - for instance if somebody sent an entire SpamAssassin report back
2475 as a large number of 550 error lines. This bug was coincidentally fixed by
2476 increasing the size of one of Exim's internal buffers (big_buffer) that
2477 happened as part of the Exiscan merge. However, to be on the safe side, I
2478 have made the code more robust (and fixed the comments that describe what
2481 53. Now that there can be additional text after "Completed" in log lines (if
2482 the queue_time_overall log selector is set), a one-byte patch to exigrep
2483 was needed to allow it to recognize "Completed" as not the last thing in
2486 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2487 patch that reportedly fixes this has been added. I am not expert enough to
2488 create a test for it. This is what the patch creator wrote:
2490 "I found a little strange behaviour of ldap code when working with
2491 Windows 2003 AD Domain, where users was placed in more than one
2492 Organization Units. When I tried to give exim partial DN, the exit code
2493 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2494 But simultaneously result of request was absolutely normal ldap result,
2495 so I produce this patch..."
2497 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2498 REFERENCE, so I have modified the code to exclude the patch when that macro
2501 55. Some experimental protocols are using DNS PTR records for new purposes. The
2502 keys for these records are domain names, not reversed IP addresses. The
2503 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2504 leaves it alone. Component reversal etc. now happens only for IP addresses.
2507 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2509 57. Double the size of the debug message buffer (to 2048) so that more of very
2510 long debug lines gets shown.
2512 58. The exicyclog utility now does better if the number of log files to keep
2513 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2515 59. Two changes related to the smtp_active_hostname option:
2517 (1) $smtp_active_hostname is now available as a variable.
2518 (2) The default for smtp_banner uses $smtp_active_hostname instead
2519 of $primary_hostname.
2521 60. The host_aton() function is supposed to be passed a string that is known
2522 to be a valid IP address. However, in the case of IPv6 addresses, it was
2523 not checking this. This is a hostage to fortune. Exim now panics and dies
2524 if the condition is not met. A case was found where this could be provoked
2525 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2526 components; fortuitously, this particular loophole had already been fixed
2527 by change 4.50/55 above.
2529 If there are any other similar loopholes, the new check in host_aton()
2530 itself should stop them being exploited. The report I received stated that
2531 data on the command line could provoke the exploit when Exim was running as
2532 exim, but did not say which command line option was involved. All I could
2533 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2534 running as the user.
2537 61. There was a buffer overflow vulnerability in the SPA authentication code
2538 (which came originally from the Samba project). I have added a test to the
2539 spa_base64_to_bits() function which I hope fixes it.
2542 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2543 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2545 63. The daemon start-up calls getloadavg() while still root for those OS that
2546 need the first call to be done as root, but it missed one case: when
2547 deliver_queue_load_max is set with deliver_drop_privilege. This is
2548 necessary for the benefit of the queue runner, because there is no re-exec
2549 when deliver_drop_privilege is set.
2551 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2552 This has been fixed.
2554 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2555 was in use, was not putting the data itself into the right store pool;
2556 consequently, it could be overwritten for a subsequent message in the same
2557 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2560 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2562 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2563 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2564 the function were treating the return as a boolean value, which happened to
2565 work because 0=false and not-0=true, but is not correct code.
2567 68. The host_aton() function was not handling scoped IPv6 addresses (those
2568 with, for example, "%eth0" on the end) correctly.
2570 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2571 negated items (that is, ~something) in unsigned ints. Some compilers
2572 apparently mutter when there is no cast.
2574 70. If an address verification called from an ACL failed, and did not produce a
2575 user-specific message (i.e. there was only a "system" message), nothing was
2576 put in $acl_verify_message. In this situation, it now puts the system
2579 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2580 session; change 4.31/43 added the unwanted input to the log line - except
2581 that it did not do this in the start of session case. It now does.
2583 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2584 This is wrong and can cause the other end to generate a synchronization
2585 error if it is another Exim or anything else that does the synchronization
2586 check. A QUIT command is no longer sent after a timeout.
2588 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2589 during host lookups.
2591 74. The defer_ok option of callout verification was not working if it was used
2592 when verifying addresses in header lines, that is, for this case:
2594 verify = header_sender/callout=defer_ok
2596 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2597 those file descriptors could be used for SMTP connections. If anything
2598 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2599 could be sent to the SMTP client, causing chaos. The daemon now opens
2600 stdin, stdout, and stderr to /dev/null when it puts itself into the
2603 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2604 log by default. The user can override this with suitable Perl magic.
2606 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2607 the log message when discard triggers, was not working for the DATA ACL or
2608 for the non-SMTP ACL.
2610 78. Error message wording change in sieve.c.
2612 79. If smtp_accept_max_per_host was set, the number of connections could be
2613 restricted to fewer than expected, because the daemon was trying to set up
2614 a new connection before checking whether the processes handling previous
2615 connections had finished. The check for completed processes is now done
2616 earlier. On busy systems, this bug wouldn't be noticed because something
2617 else would have woken the daemon, and it would have reaped the completed
2620 80. If a message was submitted locally by a user whose login name contained one
2621 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2622 It caused a spool format error. I have fixed the spool reading code. A
2623 related problem was that the "from" clause in the Received: line became
2624 illegal because of the space(s). It is now covered by ${quote_local_part.
2626 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2629 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2631 83. Patch from Sieve maintainer to fix unterminated string problem in
2632 "vacation" handling.
2634 84. Some minor changes to the Linux configuration files to help with other
2635 OS variants using glibc.
2637 85. One more patch for Sieve to update vacation handling to latest spec.
2640 ----------------------------------------------------
2641 See the note above about the 4.44 and 4.50 releases.
2642 ----------------------------------------------------
2648 1. Change 4.43/35 introduced a bug that caused file counts to be
2649 incorrectly computed when quota_filecount was set in an appendfile
2652 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2653 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2656 3. Give more explanation in the error message when the command for a transport
2657 filter fails to execute.
2659 4. There are several places where Exim runs a non-Exim command in a
2660 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2661 was being done only for the command run by the queryprogram router. It is
2662 now done for all such subprocesses. The other cases are: ${run, transport
2663 filters, and the commands run by the lmtp and pipe transports.
2665 5. Some older OS have a limit of 256 on the maximum number of file
2666 descriptors. Exim was using setrlimit() to set 1000 as a large value
2667 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2668 systems. I've change it so that if it can't get 1000, it tries for 256.
2670 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2671 was an oversight, and furthermore, ever since the addition of extra
2672 controls (e.g. 4.43/32), the checks on when to allow different forms of
2673 "control" were broken. There should now be diagnostics for all cases when a
2674 control that does not make sense is encountered.
2676 7. $recipients is now available in the predata ACL (oversight).
2678 8. Tidy the search cache before the fork to do a delivery from a message
2679 received from the command line. Otherwise the child will trigger a lookup
2680 failure and thereby defer the delivery if it tries to use (for example) a
2681 cached ldap connection that the parent has called unbind on.
2683 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2684 of $address_data from the recipient verification was clobbered by the
2685 sender verification.
2687 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2688 was its contents. (It was OK if the option was not defined at all.)
2690 11. A "Completed" log line is now written for messages that are removed from
2691 the spool by the -Mrm option.
2693 12. $host_address is now set to the target address during the checking of
2694 ignore_target_hosts.
2696 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2697 being passed; this would have caused $sender_host_name to have been used if
2698 matching the list had actually called for a host name (not very likely,
2699 since this list is usually IP addresses). A host name is now passed as
2702 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2703 code that specifies a non-restarting handler (typically sigaction() in
2704 modern systems) in an attempt to fix a rare and obscure crash bug.
2706 15. Narrowed the window for a race in the daemon that could cause it to ignore
2707 SIGCHLD signals. This is not a major problem, because they are used only to
2708 wake it up if nothing else does.
2710 16. A malformed maildirsize file could cause Exim to calculate negative values
2711 for the mailbox size or file count. Odd effects could occur as a result.
2712 The maildirsize information is now recalculated if the size or filecount
2715 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2716 support for a long time. Removed HAVE_SYS_VFS_H.
2718 18. Updated exipick to current release from John Jetmore.
2720 19. Allow an empty sender to be matched against a lookup in an address list.
2721 Previously the only cases considered were a regular expression, or an
2724 20. Exim went into a mad DNS lookup loop when doing a callout where the
2725 host was specified on the transport, if the DNS lookup yielded more than
2728 21. The RFC2047 encoding function was originally intended for short strings
2729 such as real names; it was not keeping to the 75-character limit for
2730 encoded words that the RFC imposes. It now respects the limit, and
2731 generates multiple encoded words if necessary. To be on the safe side, I
2732 have increased the buffer size for the ${rfc2047: expansion operator from
2735 22. Failure to deliver a bounce message always caused it to be frozen, even if
2736 there was an errors_to setting on the router. The errors_to setting is now
2739 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2740 canonical form (fully expanded) before being placed in
2741 $sender_host_address.
2743 24. Updated eximstats to version 1.33
2745 25. Include certificate and key file names in error message when GnuTLS fails
2746 to set them up, because the GnuTLS error message doesn't include the name
2747 of the failing file when there is a problem reading it.
2749 26. Expand error message when OpenSSL has problems setting up cert/key files.
2752 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2753 (this can affect the format of dates).
2755 28. exim_tidydb, when checking for the continued existence of a message for
2756 which it has found a message-specific retry record, was not finding
2757 messages that were in split spool directories. Consequently, it was
2758 deleting retry records that should have stayed in existence.
2760 29. eximstats updated to version 1.35
2761 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2762 1.35 - bugfix such that pie charts by volume are generated correctly
2764 30. The SPA authentication driver was not abandoning authentication and moving
2765 on to the next authenticator when an expansion was forced to fail,
2766 contradicting the general specification for all authenticators. Instead it
2767 was generating a temporary error. It now behaves as specified.
2769 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2770 (the order specifies the preference for clients). The order is now AES256,
2771 AES128, 3DES, ARCFOUR128.
2773 31. Small patch to Sieve code - explicitly set From: when generating an
2776 32. Exim crashed if a remote delivery caused a very long error message to be
2777 recorded - for instance if somebody sent an entire SpamAssassin report back
2778 as a large number of 550 error lines. This bug was coincidentally fixed by
2779 increasing the size of one of Exim's internal buffers (big_buffer) that
2780 happened as part of the Exiscan merge. However, to be on the safe side, I
2781 have made the code more robust (and fixed the comments that describe what
2784 33. Some experimental protocols are using DNS PTR records for new purposes. The
2785 keys for these records are domain names, not reversed IP addresses. The
2786 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2787 leaves it alone. Component reversal etc. now happens only for IP addresses.
2790 34. The host_aton() function is supposed to be passed a string that is known
2791 to be a valid IP address. However, in the case of IPv6 addresses, it was
2792 not checking this. This is a hostage to fortune. Exim now panics and dies
2793 if the condition is not met. A case was found where this could be provoked
2794 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2795 components; fortuitously, this particular loophole had already been fixed
2796 by change 4.50/55 or 4.44/33 above.
2798 If there are any other similar loopholes, the new check in host_aton()
2799 itself should stop them being exploited. The report I received stated that
2800 data on the command line could provoke the exploit when Exim was running as
2801 exim, but did not say which command line option was involved. All I could
2802 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2803 running as the user.
2806 35. There was a buffer overflow vulnerability in the SPA authentication code
2807 (which came originally from the Samba project). I have added a test to the
2808 spa_base64_to_bits() function which I hope fixes it.
2811 36. The daemon start-up calls getloadavg() while still root for those OS that
2812 need the first call to be done as root, but it missed one case: when
2813 deliver_queue_load_max is set with deliver_drop_privilege. This is
2814 necessary for the benefit of the queue runner, because there is no re-exec
2815 when deliver_drop_privilege is set.
2817 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2818 was in use, was not putting the data itself into the right store pool;
2819 consequently, it could be overwritten for a subsequent message in the same
2820 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2823 38. Sometimes the final signoff response after QUIT could fail to get
2824 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2825 < 0 before doing a fflush(). This bug looks as though it goes back to the
2826 introduction of TLS in release 3.20, but "sometimes" must have been rare
2827 because the tests only now provoked it.
2833 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2834 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2835 Now it returns an integer. A number of calls were still expecting a T/F
2836 return. Fortuitously, in all cases, the tests worked in OK situations,
2837 which is the norm. However, things would have gone wrong on any write
2838 failures on the smtp file descriptor. This function is used when sending
2839 messages over SMTP and also when doing verify callouts.
2841 2. When Exim is called to do synchronous delivery of a locally submitted
2842 message (the -odf or -odi options), it no longer closes stderr before doing
2845 3. Implemented the mua_wrapper option.
2847 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2849 5. Implemented the functions header_remove(), header_testname(),
2850 header_add_at_position(), and receive_remove_recipient(), and exported them
2853 6. If an ACL "warn" statement specified the addition of headers, Exim already
2854 inserted X-ACL-Warn: at the start if there was no header name. However, it
2855 was not making this test for the second and subsequent header lines if
2856 there were newlines in the string. This meant that an invalid header could
2857 be inserted if Exim was badly configured.
2859 7. Allow an ACL "warn" statement to add header lines at the start or after all
2860 the Received: headers, as well as at the end.
2862 8. Added the rcpt_4xx retry error code.
2864 9. Added postmaster_mailfrom=xxx to callout verification option.
2866 10. Added mailfrom=xxxx to the callout verification option, for verify=
2869 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2870 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2872 12. Inserted some casts to stop certain compilers warning when using pointer
2873 differences as field lengths or precisions in printf-type calls (mostly
2874 affecting debugging statements).
2876 13. Added optional readline() support for -be (dynamically loaded).
2878 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2879 same clock tick as a message's arrival, so that its received time was the
2880 same as the "first fail" time on the retry record, and that message
2881 remained on the queue past the ultimate address timeout, every queue runner
2882 would try a delivery (because it was past the ultimate address timeout) but
2883 after another failure, the ultimate address timeout, which should have then
2884 bounced the address, did not kick in. This was a "< instead of <=" error;
2885 in most cases the first failure would have been in the next clock tick
2886 after the received time, and all would be well.
2888 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2889 being recognized when the domain list was tested by the match_domain
2890 condition in an expansion string.
2892 16. Added the ${str2b64: operator.
2894 17. Exim was always calling setrlimit() to set a large limit for the number of
2895 processes, without checking whether the existing limit was already
2896 adequate. (It did check for the limit on file descriptors.) Furthermore,
2897 errors from getrlimit() and setrlimit() were being ignored. Now they are
2898 logged to the main and panic logs, but Exim does carry on, to try to do its
2899 job under whatever limits there are.
2901 18. Imported PCRE 5.0.
2903 19. Trivial typo in log message " temporarily refused connection" (the leading
2906 20. If the log selector return_path_on_delivery was set and an address was
2907 redirected to /dev/null, the delivery process crashed because it assumed
2908 that a return path would always be set for a "successful" delivery. In this
2909 case, the whole delivery is bypassed as an optimization, and therefore no
2912 21. Internal re-arrangement: the function for sending a challenge and reading
2913 a response while authentication was assuming a zero-terminated challenge
2914 string. It's now changed to take a pointer and a length, to allow for
2915 binary data in such strings.
2917 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2919 23. Exim was not respecting finduser_retries when seeking the login of the
2920 uid under which it was called; it was always trying 10 times. (The default
2921 setting of finduser_retries is zero.) Also, it was sleeping after the final
2922 failure, which is pointless.
2924 24. Implemented tls_on_connect_ports.
2926 25. Implemented acl_smtp_predata.
2928 26. If the domain in control=submission is set empty, Exim assumes that the
2929 authenticated id is a complete email address when it generates From: or
2930 Sender: header lines.
2932 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2933 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2934 chown and chgrp in /bin and hostname in /usr/bin.
2936 28. Exim was keeping the "process log" file open after each use, just as it
2937 does for the main log. This opens the possibility of it remaining open for
2938 long periods when the USR1 signal hits a daemon. Occasional processlog
2939 errors were reported, that could have been caused by this. Anyway, it seems
2940 much more sensible not to leave this file open at all, so that is what now
2943 29. The long-running daemon process does not normally write to the log once it
2944 has entered its main loop, and it closes the log before doing so. This is
2945 so that log files can straightforwardly be renamed and moved. However,
2946 there are a couple of unusual error situations where the daemon does write
2947 log entries, and I had neglected to close the log afterwards.
2949 30. The text of an SMTP error response that was received during a remote
2950 delivery was being truncated at 512 bytes. This is too short for some of
2951 the long messages that one sometimes sees. I've increased the limit to
2954 31. It is now possible to make retry rules that apply only when a message has a
2955 specific sender, in particular, an empty sender.
2957 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2958 it possible to be selective about when SMTP synchronization is enforced.
2960 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2962 32. Implemented hosts_connection_nolog.
2964 33. Added an ACL for QUIT.
2966 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2969 35. Added mailbox_size and mailbox_filecount to appendfile.
2971 36. Added control = no_multiline_responses to ACLs.
2973 37. There was a bug in the logic of the code that waits for the clock to tick
2974 in the case where the clock went backwards by a substantial amount such
2975 that the microsecond fraction of "now" was more than the microsecond
2976 fraction of "then" (but the whole seconds number was less).
2978 38. Added support for the libradius Radius client library this is found on
2979 FreeBSD (previously only the radiusclient library was supported).
2985 1. When certain lookups returned multiple values in the form name=value, the
2986 quoting of the values was not always being done properly. Specifically:
2987 (a) If the value started with a double quote, but contained no whitespace,
2989 (b) If the value contained whitespace other than a space character (i.e.
2990 tabs or newlines or carriage returns) it was not quoted.
2991 This fix has been applied to the mysql and pgsql lookups by writing a
2992 separate quoting function and calling it from the lookup code. The fix
2993 should probably also be applied to nisplus, ibase and oracle lookups, but
2994 since I cannot test any of those, I have not disturbed their existing code.
2996 2. A hit in the callout cache for a specific address caused a log line with no
2997 reason for rejecting RCPT. Now it says "Previous (cached) callout
2998 verification failure".
3000 3. There was an off-by-one bug in the queryprogram router. An over-long
3001 return line was truncated at 256 instead of 255 characters, thereby
3002 overflowing its buffer with the terminating zero. As well as fixing this, I
3003 have increased the buffer size to 1024 (and made a note to document this).
3005 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3006 when Exim is waiting for an SMTP response from a remote server, Exim
3007 restarts its select() call on the socket, thereby resetting its timeout.
3008 This is not a problem when such interrupts are rare. Somebody set up a cron
3009 job to run exiwhat every 2 minutes, which is less than the normal select()
3010 timeout (5 or 10 minutes). This meant that the select() timeout never
3011 kicked in because it was always reset. I have fixed this by comparing the
3012 time when an interrupt arrives with the time at the start of the first call
3013 to select(). If more time than the timeout has elapsed, the interrupt is
3014 treated as a timeout.
3016 5. Some internal re-factoring in preparation for the addition of Sieve
3017 extensions (by MH). In particular, the "personal" test is moved to a
3018 separate function, and given an option for scanning Cc: and Bcc: (which is
3019 not set for Exim filters).
3021 6. When Exim created an email address using the login of the caller as the
3022 local part (e.g. when creating a From: or Sender: header line), it was not
3023 quoting the local part when it contained special characters such as @.
3025 7. Installed new OpenBSD configuration files.
3027 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3028 try to make them clearer.
3030 9. Callout options, other than the timeout value, were being ignored when
3031 verifying sender addresses in header lines. For example, when using
3033 verify = header_sender/callout=no_cache
3035 the cache was (incorrectly) being used.
3037 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3038 only the Cygwin environment.
3040 11. When return_path_on_delivery was set as a log selector, if different remote
3041 addresses in the same message used different return paths and parallel
3042 remote delivery occurred, the wrong values would sometimes be logged.
3043 (Whenever a remote delivery process finished, the return path value from
3044 the most recently started remote delivery process was logged.)
3046 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3047 lines when AUTH and/or TLS are in use. This is the "received protocol"
3048 field. Exim used to use "asmtp" for authenticated SMTP, without any
3049 indication (in the protocol name) for TLS use. Now it follows the RFC and
3050 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3051 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3052 names appear in log lines as well as in Received: header lines.
3054 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3055 extensions, and comparison tests, and to fix some bugs.
3057 14. Changes to the "personal" filter test:
3059 (1) The test was buggy in that it was just doing the equivalent of
3060 "contains" tests on header lines. For example, if a user's address was
3061 anne@some.where, the "personal" test would incorrectly be true for
3063 To: susanne@some.where
3065 This test is now done by extracting each address from the header in turn,
3066 and checking the entire address. Other tests that are part of "personal"
3067 are now done using regular expressions (for example, to check local parts
3068 of addresses in From: header lines).
3070 (2) The list of non-personal local parts in From: addresses has been
3071 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3072 taken from the Sieve specification recommendations.
3074 (3) If the message contains any header line starting with "List-" it is
3075 treated as non-personal.
3077 (4) The test for "circular" in the Subject: header line has been removed
3078 because it now seems ill-conceived.
3080 15. Minor typos in src/EDITME comments corrected.
3082 16. Installed latest exipick from John Jetmore.
3084 17. If headers_add on a router specified a text string that was too long for
3085 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3086 of string_sprintf() is now avoided.
3088 18. $message_body_size was not set (it was always zero) when running the DATA
3089 ACL and the local_scan() function.
3091 19. For the "mail" command in an Exim filter, no default was being set for
3092 the once_repeat time, causing a random time value to be used if "once" was
3093 specified. (If the value happened to be <= 0, no repeat happened.) The
3094 default is now 0s, meaning "never repeat". The "vacation" command was OK
3095 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3096 (I found it when inspecting the code).
3098 20. There is now an overall timeout for performing a callout verification. It
3099 defaults to 4 times the callout timeout, which applies to individual SMTP
3100 commands during the callout. The overall timeout applies when there is more
3101 than one host that can be tried. The timeout is checked before trying the
3102 next host. This prevents very long delays if there are a large number of
3103 hosts and all are timing out (e.g. when the network connections are timing
3104 out). The value of the overall timeout can be changed by specifying an
3105 additional sub-option for "callout", called "maxwait". For example:
3107 verify = sender/callout=5s,maxwait=20s
3109 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3110 to the end before writing, but this should make it even safer).
3112 22. Exim was forgetting that it had advertised PIPELINING for the second and
3113 subsequent messages on an SMTP connection. It was also not resetting its
3114 memory on STARTTLS and an internal HELO.
3116 23. When Exim logs an SMTP synchronization error within a session, it now
3117 records whether PIPELINING has been advertised or not.
3119 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3120 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3121 rather than long int.
3123 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3125 26. Added the never_mail option to autoreply.
3131 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3132 crash if the getsockname() call failed; this can happen if a connection is
3133 closed very soon after it is established. The problem was simply in the
3134 order in which certain operations were done, causing Exim to try to write
3135 to the SMTP stream before it had set up the file descriptor. The bug has
3136 been fixed by making things happen in the correct order.
3142 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3143 before the connection was closed, thus losing the rejection response.
3145 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3146 some early Solaris releases, but causes trouble in current releases where
3147 socklen_t is defined.
3149 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3152 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3155 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3156 that releasing the top of it at the end releases what was used for sub-
3157 expansions (unless the block got too big). However, discard this block if
3158 the first thing is a variable or header, so that we can use its block when
3159 it is dynamic (useful for very large $message_headers, for example).
3161 6. Lookups now cache *every* query, not just the most recent. A new, separate
3162 store pool is used for this. It can be recovered when all lookup caches are
3163 flushed. Lookups now release memory at the end of their result strings.
3164 This has involved some general refactoring of the lookup sources.
3166 7. Some code has been added to the store_xxx() functions to reduce the amount
3167 of flapping under certain conditions.
3169 8. log_incoming_interface used to affect only the <= reception log lines. Now
3170 it causes the local interface and port to be added to several more SMTP log
3171 lines, for example "SMTP connection from", and rejection lines.
3173 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3175 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3177 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3178 could be overwritten at the end of the current message (or the start of a
3179 new message if it was set in a HELO ACL). The value is now preserved for
3180 the duration of the SMTP connection.
3182 12. If a transport had a headers_rewrite setting, and a matching header line
3183 contained an unqualified address, that address was qualified, even if it
3184 did not match any rewriting rules. The underlying bug was that the values
3185 of the flags that permit the existence of unqualified sender and recipient
3186 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3187 non-local messages, and by -bnq for local messages) were not being
3188 preserved with the message after it was received.
3190 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3191 "next input=" as part of the text comprising the host identity instead of
3192 the correct text. The code was using the same buffer for two different
3193 strings. However, depending on which order the printing function evaluated
3194 its arguments, the bug did not always show up. Under Linux, for example, my
3195 test suite worked just fine.
3197 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3198 doesn't work with some older versions of Perl. It has been changed to "my",
3199 which in any case is probably the better facility to use.
3201 15. A really picky compiler found some instances of statements for creating
3202 error messages that either had too many or two few arguments for the format
3205 16. The size of the buffer for calls to the DNS resolver has been increased
3206 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3207 for addresses that have a lot of PTR records. This alleviates a problem; it
3208 does not fully solve it.
3210 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3211 buffer now truncates the list and logs the incident, which is the same
3212 action as happens when Exim is looking up a host name and its aliases.
3213 Previously in this situation something unpredictable would happen;
3214 sometimes it was "internal error: store_reset failed".
3216 18. If a server dropped the connection unexpectedly when an Exim client was
3217 using GnuTLS and trying to read a response, the client delivery process
3218 crashed while trying to generate an error log message.
3220 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3221 string, for example:
3223 warn message = H1: something\nH2: something
3225 the text was added as a single header line from Exim's point of view
3226 though it ended up OK in the delivered message. However, searching for the
3227 second and subsequent header lines using $h_h2: did not work. This has been
3228 fixed. Similarly, if a system filter added multiple headers in this way,
3229 the routers could not see them.
3231 20. Expanded the error message when iplsearch is called with an invalid key to
3232 suggest using net-iplsearch in a host list.
3234 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3235 ACLs are no longer actually imposed (and a message to that effect is
3238 22. If a "gecos" field in a passwd entry contained escaped characters, in
3239 particular, if it contained a \" sequence, Exim got it wrong when building
3240 a From: or a Sender: header from that name. A second bug also caused
3241 incorrect handling when an unquoted " was present following a character
3242 that needed quoting.
3244 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3245 was not being matched caselessly.
3247 24. Arranged for all hyphens in the exim.8 source to be escaped with
3250 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3251 properly. Recipient callout cache records were still being keyed to include
3252 the sender, even when use_sender was set false. This led to far more
3253 callouts that were necessary. The sender is no longer included in the key
3254 when use_sender is false.
3256 26. Added "control = submission" modifier to ACLs.
3258 27. Added the ${base62d: operator to decode base 62 numbers.
3260 28. dnsdb lookups can now access SRV records.
3262 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3263 the configuration file.
3265 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3266 (-v) mode. This makes the output for a verbose queue run more intelligible.
3268 31. Added a use_postmaster feature to recipient callouts.
3270 32. Added the $body_zerocount variable, containing the number of binary zero
3271 bytes in the message body.
3273 33. The time of last modification of the "new" subdirectory is now used as the
3274 "mailbox time last read" when there is a quota error for a maildir
3277 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3279 35. Added +ignore_unknown as a special item in host lists.
3281 36. Code for decoding IPv6 addresses in host lists is now included, even if
3282 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3283 address was recognized as an IP address, but was then not correctly decoded
3284 into binary, causing unexpected and incorrect effects when compared with
3291 1. Very minor rewording of debugging text in manualroute to say "list of
3292 hosts" instead of "hostlist".
3294 2. If verify=header_syntax was set, and a header line with an unqualified
3295 address (no domain) and a large number of spaces between the end of the
3296 name and the colon was received, the reception process suffered a buffer
3297 overflow, and (when I tested it) crashed. This was caused by some obsolete
3298 code that should have been removed. The fix is to remove it!
3300 3. When running in the test harness, delay a bit after writing a bounce
3301 message to get a bit more predictability in the log output.
3303 4. Added a call to search_tidyup() just before forking a reception process. In
3304 theory, someone could use a lookup in the expansion of smtp_accept_max_
3305 per_host which, without the tidyup, could leave open a database connection.
3307 5. Added the variables $recipient_data and $sender_data which get set from a
3308 lookup success in an ACL "recipients" or "senders" condition, or a router
3309 "senders" option, similar to $domain_data and $local_part_data.
3311 6. Moved the writing of debug_print from before to after the "senders" test
3314 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3315 problems for message scanning, either using a data ACL, or using
3316 local_scan() because the Received: header was not generated till after they
3317 were called (in order to set the time as the time of reception completion).
3318 I have revised the way this works. The header is now generated after the
3319 body is received, but before the ACL or local_scan() are called. After they
3320 are run, the timestamp in the header is updated.
3326 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3327 before starting a queue runner without re-exec. This happened only when
3328 deliver_drop_privilege was set or when the Exim user was set to root. The
3329 effect of the bug was that timeouts during subsequent deliveries caused
3330 crashes instead of being properly handled. The handler is now left at its
3331 default (and expected) setting.
3333 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3334 message, again when deliver_drop_privilege is set or Exim is run as root.
3335 The bug described in (1) was not present in this case, but the tidying up
3336 of the other signals was missing. I have made the two cases consistent.
3338 3. The ignore_target_hosts setting on a manualroute router was being ignored
3339 for hosts that were looked up using the /MX notation.
3341 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3344 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3345 operated on the sender address. After changing the $sender_address to <>
3346 for the sender address verify, Exim was re-instated it as the original
3347 (before rewriting) address, but remembering that it had rewritten it, so it
3348 wasn't rewriting it again. This bug also had the effect of breaking the
3349 sender address verification caching when the sender address was rewritten.
3351 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3352 This has been changed so that if the ip literal address matches
3353 ignore_target_hosts, the router declines.
3355 7. Added expansion conditions match_domain, match_address, and match_local_
3356 part (NOT match_host).
3358 8. The placeholder for the Received: header didn't have a length field set.
3360 9. Added code to Exim itself and to exim_lock to test for a specific race
3361 condition that could lead to file corruption when using MBX delivery. The
3362 issue is with the lockfile that is created in /tmp. If this file is removed
3363 after a process has opened it but before that process has acquired a lock,
3364 there is the potential for a second process to recreate the file and also
3365 acquire a lock. This could lead to two Exim processes writing to the file
3366 at the same time. The added code performs the same test as UW imapd; it
3367 checks after acquiring the lock that its file descriptor still refers to
3368 the same named file.
3370 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3371 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3374 11. Added the smtp_active_hostname option. If used, this will typically be made
3375 to depend on the incoming interface address. Because $interface_address is
3376 not set up until the daemon has forked a reception process, error responses
3377 that can happen earlier (such as "too many connections") no longer contain
3380 12. If an expansion in a condition on a "warn" statement fails because a lookup
3381 defers, the "warn" statement is abandoned, and the next ACL statement is
3382 processed. Previously this caused the whole ACL to be aborted.
3384 13. Added the iplsearch lookup type.
3386 14. Added ident_timeout as a log selector.
3388 15. Added tls_certificate_verified as a log selector.
3390 16. Added a global option tls_require_ciphers (compare the smtp transport
3391 option of the same name). This controls incoming TLS connections.
3393 17. I finally figured out how to make tls_require_ciphers do a similar thing
3394 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3395 before starting the TLS session.
3397 18. Tabs are now shown as \t in -bP output.
3399 19. If the log selector return_path_on_delivery was set, Exim crashed when
3400 bouncing a message because it had too many Received: header lines.
3402 20. If two routers both had headers_remove settings, and the first one included
3403 a superfluous trailing colon, the final name in the first list and the
3404 first name in the second list were incorrectly joined into one item (with a
3405 colon in the middle).
3411 1. Added -C and -D options to the exinext utility, mainly to make it easier
3412 to include in the automated testing, but these could be helpful when
3413 multiple configurations are in use.
3415 2. The exinext utility was not formatting the output nicely when there was
3416 an alternate port involved in the retry record key, nor when there was a
3417 message id as well (for retries that were specific to a specific message
3418 and a specific host). It was also confused by IPv6 addresses, because of
3419 the additional colons they contain. I have fixed the IPv4 problem, and
3420 patched it up to do a reasonable job for IPv6.
3422 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3423 delivery, the log line now contains "pipelined" if PIPELINING was used.
3425 4. An SMTP transport process used to panic and die if the bind() call to set
3426 an explicit outgoing interface failed. This has been changed; it is now
3427 treated in the same way as a connect() failure.
3429 5. A reference to $sender_host_name in the part of a conditional expansion
3430 that was being skipped was still causing a DNS lookup. This no longer
3433 6. The def: expansion condition was not recognizing references to header lines
3434 that used bh_ and bheader_.
3436 7. Added the _cache feature to named lists.
3438 8. The code for checking quota_filecount in the appendfile transport was
3439 allowing one more file than it should have been.
3441 9. For compatibility with Sendmail, the command line option
3449 and sets the incoming protocol and host name (for trusted callers). The
3450 host name and its colon can be omitted when only the protocol is to be set.
3451 Note the Exim already has two private options, -pd and -ps, that refer to
3452 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3453 "s", but I don't think that's a major issue.
3455 10. A number of refactoring changes to the code, none of which should affect
3458 (a) The number of logging options was getting close to filling up the
3459 32-bit word that was used as a bit map. I have split them into two classes:
3460 those that are passed in the argument to log_write(), and those that are
3461 only ever tested independently outside of that function. These are now in
3462 separate 32-bit words, so there is plenty of room for expansion again.
3463 There is no change in the user interface or the logging behaviour.
3465 (b) When building, for example, log lines, the code previously used a
3466 macro that called string_cat() twice, in order to add two strings. This is
3467 not really sufficiently general. Furthermore, there was one instance where
3468 it was actually wrong because one of the argument was used twice, and in
3469 one call a function was used. (As it happened, calling the function twice
3470 did not affect the overall behaviour.) The macro has been replaced by a
3471 function that can join an arbitrary number of extra strings onto a growing
3474 (c) The code for expansion conditions now uses a table and a binary chop
3475 instead of a serial search (which was left over from when there were very
3476 few conditions). Also, it now recognizes conditions like "pam" even when
3477 the relevant support is not compiled in: a suitably worded error message is
3478 given if an attempt is made to use such a condition.
3480 11. Added ${time_interval:xxxxx}.
3482 12. A bug was causing one of the ddress fields not to be passed back correctly
3483 from remote delivery subprocesses. The field in question was not being
3484 subsequently used, so this caused to problems in practice.
3486 13. Added new log selectors queue_time and deliver_time.
3488 14. Might have fixed a bug in maildirsizefile handling that threw up
3489 "unexpected character" debug warnings, and recalculated the data
3490 unnecessarily. In any case, I expanded the warning message to give more
3493 15. Added the message "Restricted characters in address" to the statements in
3494 the default ACL that block characters like @ and % in local parts.
3496 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3497 Three changes have been made:
3499 (a) There was a serious bug; a negative response to MAIL caused the whole
3500 recipient domain to be cached as invalid, thereby blocking all messages
3501 to all local parts at the same domain, from all senders. This bug has
3502 been fixed. The domain is no longer cached after a negative response to
3503 MAIL if the sender used is not empty.
3505 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3508 (c) A new callout option, "use_sender" has been added for people who want
3509 the modified behaviour.
3515 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3518 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3519 indeed breaks things for older releases.
3521 3. Added additional logging to the case where there is a problem reading data
3522 from a filter that is running in a subprocess using a pipe, in order to
3523 try to track down a specific problem.
3525 4. Testing facility fudge: when running in the test harness and attempting
3526 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3527 getting "No route to host". Convert this to a timeout.
3529 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3532 6. Some OS don't have socklen_t but use size_t instead. This affects the
3533 fifth argument of getsockopt() amongst other things. This is now
3534 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3535 can be set for individual OS. I have set it for SunOS5, OSF1, and
3536 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3537 some earlier ones do not.
3539 7. Change 4.30/15 was not doing the test caselessly.
3541 8. The standard form for an IPv6 address literal was being rejected by address
3542 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3543 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3544 this, as well as the form without the "IPv6" on the front (but only when
3545 address literals are enabled, of course).
3547 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3549 10. Exim crashed if a message with an empty sender address specified by -f
3550 encountered a router with an errors_to setting. This could be provoked only
3551 by a command such as
3555 where an empty string was supplied; "<>" did not hit this bug.
3557 11. Installed PCRE release 4.5.
3559 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3560 remained set. It is now erased.
3562 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3563 times from message ids (which are base 36 rather than the normal 62).
3565 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3566 were being counted as actual protocol errors, and logged if the log
3567 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3568 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3569 and DATA following a set of rejected RCPTs do not count as protocol errors.
3570 In other words, Exim assumes they were pipelined, though this may not
3571 actually be the case. Of course, in all cases the client gets an
3572 appropriate error code.
3574 15. If a lookup fails in an ACL condition, a message about the failure may
3575 be available; it is used if testing the ACL cannot continue, because most
3576 such messages specify what the cause of the deferral is. However, some
3577 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3578 that caused an old message to be retained and used if a later statement
3579 caused a defer, replacing the real cause of the deferral.
3581 16. If an IP address had so many PTR records that the DNS lookup buffer
3582 was not large enough to hold them, Exim could crash while trying to process
3583 the truncated data. It now detects and logs this case.
3585 17. Further to 4.21/58, another change has been made: if (and only if) the
3586 first line of a message (the first header line) ends with CRLF, a bare LF
3587 in a subsequent header line has a space inserted after it, so as not to
3588 terminate the header.
3590 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3591 unnecessarily, used atoi() instead of strtol(), and didn't check the
3592 termination when getting file sizes from file names by regex.
3594 19. Completely re-implemented the support for maildirsize files, in the light
3595 of a number of problems with the previous contributed implementation
3596 (4.30/29). In particular:
3598 . If the quota is zero, the maildirsize file is maintained, but no quota is
3601 . If the maildir directory does not exist, it is created before any attempt
3602 to write a maildirsize file.
3604 . The quota value in the file is just a cache; if the quota is changed in
3605 the transport, the new value overrides.
3607 . A regular expression is available for excluding directories from the
3610 20. The autoreply transport checks the characters in options that define the
3611 message's headers; it allows continued headers, but it was checking with
3612 isspace() after an embedded newline instead of explicitly looking for a
3615 21. If all the "regular" hosts to which an address was routed had passed their
3616 expiry times, and had not reached their retry times, the address was
3617 bounced, even if fallback hosts were defined. Now Exim should go on to try
3620 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3621 equivalent code in the SMTP transport. Some hosts send humungous responses
3622 to HELO/EHLO, more than 1024 it seems.
3624 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3625 gives compiler warnings in some environments. I've now done it "properly",
3628 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3629 (because of problems with the built-in one) was declared to return uschar *
3630 instead of char *, causing compiler failure.
3632 25. Fixed a file descriptor leak when processing alias/forward files.
3634 26. Fixed a minor format string issue in dbfn.c.
3636 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3638 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3639 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3642 29. When a sender address is verified, it is cached, to save repeating the test
3643 when there is more than one recipient in a message. However, when the
3644 verification involves a callout, it is possible for different callout
3645 options to be set for different recipients. It is too complicated to keep
3646 track of this in the cache, so now Exim always runs a verification when a
3647 callout is required, relying on the callout cache for the optimization.
3648 The overhead is duplication of the address routing, but this should not be
3651 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3652 to be verified with callout=postmaster, and the main callout worked but the
3653 postmaster check failed, the verification correctly failed. However, if a
3654 subsequent RCPT command asked for sender verification *without* the
3655 postmaster check, incorrect caching caused this verification also to fail,
3658 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3659 it was not caching the DNS options (qualify_single, search_parents) that
3660 were used when the lookup failed. A subsequent lookup with different
3661 options therefore always gave the same answer, though there were cases
3662 where it should not have. (Example: a "domains = !$mx_any" option on a
3663 dnslookup router: the "domains" option is always processed without any
3664 widening, but the router might have qualify_single set.) Now Exim uses the
3665 cached value only when the same options are set.
3667 32. Added John Jetmore's "exipick" utility to the distribution.
3669 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3670 than a timeout (e.g. a certificate is required, and is not provided), an
3671 Exim server now closes the connection immediately. Previously it waited for
3672 the client to close - but if the client is SSL, it seems that they each
3673 wait for each other, leading to a delay before one of them times out.
3675 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3676 maintained 0.8.x compatibility because I don't think many are using it, and
3677 it is clearly obsolete.
3679 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3682 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3683 client certificate was expired. A simple patch fixes this, though I don't
3684 understand the full logic of why the verify callback is called multiple
3687 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3688 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3689 which causes problems with some clients (such as the Certicom SSL Plus
3690 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3691 disables the coutermeasure allowing Eudora to connect."
3693 38. Exim was not checking that a write() to a log file succeeded. This could
3694 lead to Bad Things if a log got too big, in particular if it hit a file
3695 size limit. Exim now panics and dies if it cannot write to a log file, just
3696 as it does if it cannot open a log file.
3698 39. Modified OS/Makefile-Linux so that it now contains
3700 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3702 The two -D definitions ensure that Exim is compiled with large file
3703 support, which makes it possible to handle log files that are bigger than
3706 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3707 instance) a domain was checked against a named list that involved a lookup,
3708 causing $domain_data to be set, then another domain was checked against the
3709 same list, then the first domain was re-checked, the value of $domain_data
3710 after the final check could be wrong. In particular, if the second check
3711 failed, it could be set empty. This bug probably also applied to
3714 41. The strip_trailing_dot option was not being applied to the address given
3715 with the -f command-line option.
3717 42. The code for reading a message's header from the spool was incrementing
3718 $received_count, but never initializing it. This meant that the value was
3719 incorrect (doubled) while delivering a message in the same process in which
3720 it was received. In the most common configuration of Exim, this never
3721 happens - a fresh exec is done - but it can happen when
3722 deliver_drop_privilege is set.
3724 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3725 it now includes up to 150 characters of the unexpected data in the log
3728 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3729 and building data strings. The size of both of these buffers was 10 000
3730 bytes - far larger than anybody would *ever* want, thought I. Needless to
3731 say, somebody hit the limit. I have increased the maximum line length to
3732 20 000 and the maximum data length of concatenated lines to 100 000. I have
3733 also fixed two bugs, because there was no checking on these buffers. Tsk,
3734 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3735 buffer is too small.
3737 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3738 lsearch lookups. Now it does.
3740 46. When parsing a route_list item in a manualroute router, a fixed-length
3741 buffer was used for the list of hosts. I made this 1024 bytes long,
3742 thinking that nobody would ever have a list of hosts that long. Wrong.
3743 Somebody had a whole pile of complicated expansion conditions, and the
3744 string was silently truncated, leading to an expansion error. It turns out
3745 that it is easier to change to an unlimited length (owing to other changes
3746 that have happened since this code was originally written) than to build
3747 structure for giving a limitation error. The length of the item that
3748 expands into the list of hosts is now unlimited.
3750 47. The lsearch lookup could not handle data where the length of text line was
3751 more than 4095 characters. Such lines were truncated, leading to shortened
3752 data being returned. It should now handle lines of any length.
3754 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3755 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3758 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3759 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3760 no longer happens in comment lines. A list of replacements is now placed
3761 at the head of all of the source files, except those whose only change is
3762 to replace PERL_COMMAND in the very first #! line.
3764 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3765 messages on the queue, with a bottom-up merge sort, using code contributed
3766 by Michael Haardt. This should make operations like -bp somewhat faster on
3767 large queues. It won't affect queue runners, except when queue_run_in_order
3770 51. Installed eximstats 1.31 in the distribution.
3772 52. Added support for SRV lookups to the dnslookup router.
3774 53. If an ACL referred to $message_body or $message_body_end, the value was not
3775 reset for any messages that followed in the same SMTP session.
3777 54. The store-handling optimization for building very long strings was not
3778 differentiating between the different store pools. I don't think this
3779 actually made any difference in practice, but I've tidied it.
3781 55. While running the routers to verify a sender address, $sender_address
3782 was still set to the sender address. This is wrong, because when routing to
3783 send a bounce to the sender, it would be empty. Therefore, I have changed
3784 it so that, while verifying a sender address, $sender_address is set to <>.
3785 (There is no change to what happens when verifying a recipient address.)
3787 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3788 target A or AAAA records (if not already returned) without resetting the
3789 qualify_single or search_parents options of the DNS resolver. These are
3790 inappropriate in this case because the targets of MX and SRV records must
3791 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3792 target that, when qualified, matched something in the local domain. These
3793 two options are now turned off when doing these lookups.
3795 57. It seems that at least some releases of Reiserfs (which does not have the
3796 concept of a fixed number of inodes) returns zero and not -1 for the
3797 number of available inodes. This interacted badly with check_spool_inodes,
3798 which assumed that -1 was the "no such thing" setting. What I have done is
3799 to check that the total number of inodes is greater than zero before doing
3800 the test of how many are available.
3802 58. When a "warn" ACL statement has a log_message modifier, the message is
3803 remembered, and not repeated. This is to avoid a lot of repetition when a
3804 message has many recipients that cause the same warning to be written.
3805 Howewer, Exim was preserving the list of already written lines for an
3806 entire SMTP session, which doesn't seem right. The memory is now reset if a
3807 new message is started.
3809 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3810 debugging output unless log_rewrite was also set.
3812 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3813 of (int)(handle) when we know that handle contains (void *)(-1).
3815 61. The Exim daemon panic-logs an error return when it closes the incoming
3816 connection. However "connection reset by peer" seems to be common, and
3817 isn't really an error worthy of noting specially, so that particular error
3820 62. When Exim is trying to find all the local interfaces, it used to panic and
3821 die if the ioctl to get the interface flags failed. However, it seems that
3822 on at least one OS (Solaris 9) it is possible to have an interface that is
3823 included in the list of interfaces, but for which you get a failure error
3824 for this call. This happens when the interface is not "plumbed" into a
3825 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3826 failure of the "get flags" call assumes that the interface is down.
3828 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3829 makes life easier for people who are doing arithmetic on fields extracted
3830 from dates, where you often get leading zeros that should not be
3831 interpreted as octal.
3833 64. Added qualify_domain to the redirect router, to override the global
3836 65. If a pathologically long header line contained very many addresses (the
3837 report of this problem mentioned 10 000) and each of them was rewritten,
3838 Exim could use up a very large amount of memory. (It kept on making new
3839 copies of the header line as it rewrote, and never released the old ones.)
3840 At the expense of a bit more processing, the header rewriting function has
3841 been changed so that it no longer eats memory in this way.
3843 66. The generation of the Received: header has been moved from the time that a
3844 message starts to be received, to the time that it finishes. The timestamp
3845 in the Received: header should now be very close to that of the <= log
3846 line. There are two side-effects of this change:
3848 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3849 logged header lines no longer include the local Received: line, because
3850 it has not yet been created. The same applies to a copy of the message
3851 that is returned to a non-SMTP sender when a message is rejected.
3853 (b) When a filter file is tested using -bf, no additional Received: header
3854 is added to the test message. After some thought, I decided that this
3857 This change does not affect the value of $received_for. It is still set
3858 after address rewriting, but before local_scan() is called.
3860 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3862 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3863 gave an unhelpful panic error message, and a defer error. I have managed to
3864 change this behaviour so that it now rejects any supplied certificate,
3865 which seems right, as the list of acceptable certificates is empty.
3867 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3868 gave an unhelpful defer error. I have not managed to make this reject any
3869 supplied certificates, but the error message it gives is "no certificate
3870 supplied", which is not helpful.
3872 70. exigrep's output now also includes lines that are not associated with any
3873 message, but which match the given pattern. Implemented by a patch from
3874 Martin Sluka, which also tidied up the Perl a bit.
3876 71. Recipient callout verification, like sender verification, was using <> in
3877 the MAIL FROM command. This isn't really the right thing, since the actual
3878 sender may affect whether the remote host accepts the recipient or not. I
3879 have changed it to use the actual sender in the callout; this means that
3880 the cache record is now keyed on a recipient/sender pair, not just the
3881 recipient address. There doesn't seem to be a real danger of callout loops,
3882 since a callout by the remote host to check the sender would use <>.
3883 [SEE ABOVE: changed after hitting problems.]
3885 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3886 temporary errors. However, in the case of such a code being given after
3887 the end of a data transmission (i.e. after ".") Exim was failing to write
3888 a retry record for the message. (Yes, there was some broken host that was
3889 actually sending 8xx at this point.)
3891 73. An unknown lookup type in a host list could cause Exim to panic-die when
3892 the list was checked. (An example that provoked this was putting <; in the
3893 middle of a list instead of at the start.) If this happened during a DATA
3894 ACL check, a -D file could be left lying around. This kind of configuration
3895 error no longer causes Exim to die; instead it causes a defer errror. The
3896 incident is still logged to the main and panic logs.
3898 74. Buglet left over from Exim 3 conversion. The message "too many messages
3899 in one connection" was written to the rejectlog but not the mainlog, except
3900 when address rewriting (yes!) was being logged.
3902 75. Added write_rejectlog option.
3904 76. When a system filter was run not as root (that is, when system_filter_user
3905 was set), the values of the $n variables were not being returned to the
3906 main process; thus, they were not subsequently available in the $sn
3909 77. Added +return_path_on_delivery log selector.
3911 78. A connection timeout was being treated differently from recipients deferred
3912 when testing hosts_max_try with a message that was older than the host's
3913 retry timeout. (The host should not be counted, thus allowing all hosts to
3914 be tried at least once before bouncing.) This may have been the cause of an
3915 occasionally reported bug whereby a message would remain on the queue
3916 longer than the retry timeout, but would be bounced if a delivery was
3917 forced. I say "may" because I never totally pinned down the problem;
3918 setting up timeout/retry tests is difficult. See also the next item.
3920 79. The ultimate address timeout was not being applied to errors that involved
3921 a combination of host plus message (for example, a timeout on a MAIL
3922 command). When an address resolved to a number of possible hosts, and they
3923 were not all tried for each delivery (e.g. because of hosts_max_try), a
3924 message could remain on the queue longer than the retry timeout.
3926 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3929 81. Fixed an obscure SMTP outgoing bug which required at least the following
3930 conditions: (a) there was another message waiting for the same server;
3931 (b) the server returned 5xx to all RCPT commands in the first message so
3932 that the message was not completed; (c) the server dropped the connection
3933 or gave a negative response to the RSET that Exim sends to abort the
3934 transaction. The observed case was a dropped connection after DATA that had
3935 been sent in pipelining mode. That is, the server had advertised PIPELINING
3936 but was not implementing it correctly. The effect of the bug was incorrect
3937 behaviour, such as trying another host, and this could lead to a crash.
3943 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3944 and daemon.c were passed as pointers to ints; they should have been
3945 pointers to socklen_t variables (which are typically unsigned ints).
3947 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3950 3. Fixed a really odd bug that affected only the testing scheme; patching a
3951 certain fixed string in the binary changed the value of another string that
3952 happened to be identical to the end of the original first string.
3954 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3955 name", it returns that address as the IP address. On some operating
3956 systems (e.g. Solaris), it also passes back the IP address string as the
3957 "host name". However, on others (e.g. Linux), it passes back an empty
3958 string. Exim wasn't checking for this, and was changing the host name to an
3959 empty string, assuming it had been canonicized.
3961 5. Although rare, it is permitted to have more than one PTR record for a given
3962 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3963 all the names associated with an address, because they do in Solaris.
3964 However, it seems that they do not in Linux for data that comes from the
3965 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3966 I found this out when I moved to a new Linux workstation and tried to run
3967 the Exim test suite.
3969 To get round this problem I have changed the code so that it now does its
3970 own call to the DNS to look up PTR records when searching for a host name.
3971 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3972 addresses that are only in /etc/hosts are still found.
3974 This behaviour is, however, controlled by an option called host_lookup_
3975 order, which defaults to "bydns:byaddr". If people want to use the other
3976 order, or indeed, just use one or the other means of lookup, they can
3977 specify it in this variable.
3979 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3980 some operating systems, this comes back from gethostbyaddr() as an empty
3981 string, and this is what Exim used to test for. However, it seems that in
3982 other systems, "." is yielded. Exim now tests for this case too.
3984 7. The values of check_spool_space and check_log_space are now held internally
3985 as a number of kilobytes instead of an absolute number of bytes. If a
3986 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3987 kilobyte. This means that much larger values can be stored.
3989 8. Exim monitor: an attempt to get the action menu when not actually pointing
3990 at a message produces an empty menu entitled "No message selected". This
3991 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3992 no entries in it ("Shell widget menu has zero width and/or height"). So I
3993 have added a single, blank menu entry in this case.
3995 9. Added ${quote_local_part.
3997 10. MIME decoding is now applied to the contents of Subject: header lines when
4000 11. Now that a reference to $sender_host_address automatically causes a reverse
4001 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4002 host lookup before query-style lookups in lists that might use this
4003 variable. This has therefore been abolished, and the "net-" prefix is no
4004 longer necessary for query-style lookups.
4006 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4007 have been a typo for LFLAGS, so it has been changed.
4009 13. The install script calls Exim with "-C /dev/null" in order to find the
4010 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4011 to be output. Howeve, since Exim outputs its version number before the
4012 error, it didn't break the script. It just looked ugly. I fixed this by
4013 always allowing "-C /dev/null" if the caller is root.
4015 14. Ignore overlarge ACL variable number when reading spool file - insurance
4016 against a later release with more variables having written the file.
4018 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4019 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4020 this, as well as the form without the "IPv6" on the front.
4022 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4023 OS/Makefile-Darwin file.
4025 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4026 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4028 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4029 variable. However, some OS use ints for inodes. I've added cast to long int
4030 to get rid of the compiler warning.
4032 19. I had forgotten to lock out "/../" in configuration file names when
4033 ALT_CONFIG_PREFIX was set.
4035 20. Routers used for verification do not need to specify transports. However,
4036 if such a router generated a host list, and callout was configured, Exim
4037 crashed, because it could not find a port number from the (non-existent)
4038 transport. It now assumes port 25 in this circumstance.
4040 21. Added the -t option to exigrep.
4042 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4043 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4044 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4045 them is set, LOOKUP_LSEARCH is forced.
4047 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4049 24. Added sender and host information to the "rejected by local_scan()" log
4050 line; previously there was no indication of these.
4052 25. Added .include_if_exists.
4054 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4055 Linux. It turns out that not all file systems support this. Apparently some
4056 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4057 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4058 not supported on the file descriptor, is now ignored when Exim is trying to
4059 sync a directory. This applies only to Linux.
4061 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4063 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4064 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4067 29. Added support for maildirsize files from supplied patch (modified a bit).
4069 30. The use of :fail: followed by an empty string could lead Exim to respond to
4070 sender verification failures with (e.g.):
4072 550 Verification failed for <xxx>
4073 550 Sender verify failed
4075 where the first response line was missing the '-' that indicates it is not
4076 the final line of the response.
4078 31. The loop for finding the name of the user that called Exim had a hardwired
4079 limit of 10; it now uses the value of finduser_retries, which is used for
4080 all other user lookups.
4082 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4085 33. Exim was neglecting to zero errno before one call of strtol() when
4086 expanding a string and expecting an integer value. On some systems this
4087 resulted in spurious "integer overflow" errors. Also, it was casting the
4088 result into an int without checking.
4090 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4091 did not work. The code looks as if it has *never* worked, though it appears
4092 to have been documented since at least releast 1.62. I have made it work.
4094 35. The "timeout_DNS" error in retry rules, also documented since at least
4095 1.62, also never worked. As it isn't clear exactly what this means, and
4096 clearly it isn't a major issue, I have abolished the feature by treating it
4097 as "timeout", and writing a warning to the main and panic logs.
4099 36. The display of retry rules for -brt wasn't always showing the error code
4102 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4103 timeout_connect_A, timeout_connect_MX.
4105 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4106 to the empty sender.
4108 39. The daemon was not analysing the content of -oX till after it had closed
4109 stderr and disconnected from the controlling terminal. This meant that any
4110 syntax errors were only noted on the panic log, and the return code from
4111 the command was 0. By re-arranging the code a little, I've made the
4112 decoding happen first, so such errors now appear on stderr, and the return
4113 code is 1. However, the actual setting up of the sockets still happens in
4114 the disconnected process, so errors there are still only recorded on the
4117 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4118 connections (as happens on some IP stacks) was logged at start up time as
4119 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4120 it from "IPv6 and IPv4", which means that two separate sockets are being
4123 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4124 says whether AF_INET or AF_INET6 was passed as an argument.
4126 42. Exiwhat output was messed up when time zones were included in log
4129 43. Exiwhat now gives more information about the daemon's listening ports,
4130 and whether -tls-on-connect was used.
4132 44. The "port" option of the smtp transport is now expanded.
4134 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4135 silently ignored. Now an error message is written to the main and panic
4138 46. There's a new ACL modifier called "logwrite" which writes to a log file
4139 as soon as it is encountered.
4141 47. Added $local_user_uid and $local_user_gid at routing time.
4143 48. Exim crashed when trying to verify a sender address that was being
4146 49. Exim was recognizing only a space character after ".include". It now also
4147 recognizes a tab character.
4149 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4150 extracting the relevant information from the specification. The man page no
4151 longer contains scrambled data for the -d option, and I've added a section
4152 at the front about calling Exim under different names.
4154 51. Added "extra_headers" argument to the "mail" command in filter files.
4156 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4159 53. Installed eximstats 1.29.
4161 54. Added transport_filter_timeout as a generic transport option.
4163 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4164 Cc: header lines. This was required by RFC 822, but it not required by RFC
4167 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4168 incoming messages that did not have them. Now it does so only if the
4169 message originates locally, that is, if there is no associated remote host
4170 address. When Resent- header lines are present, this applies to the Resent-
4171 lines rather than the non-Resent- lines.
4173 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4174 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4176 58. Messages for configuration errors now include the name of the main
4177 configuration file - useful now that there may be more than one file in a
4178 list (.included file names were always shown).
4180 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4181 for those rare installations that do not start the daemon as root or run it
4182 setuid root. I've cut out the call to initgroups() if the daemon is not
4185 60. The Exim user and group can now be bound into the binary as text strings
4186 that are looked up at the start of Exim's processing.
4188 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4190 62. Added $mailstore_basename variable.
4192 63. Installed patch to sieve.c from Michael Haardt.
4194 64. When Exim failed to open the panic log after failing to open the main log,
4195 the original message it was trying to log was written to stderr and debug
4196 output, but if they were not available (the usual case in production), it
4197 was lost. Now it is written to syslog before the two lines that record the
4198 failures to open the logs.
4200 65. Users' Exim filters run in subprocesses under the user's uid. It is
4201 possible for a "deliver" command or an alias in a "personal" command to
4202 provoke an address rewrite. If logging of address rewriting is configured,
4203 this fails because the process is not running as root or exim. There may be
4204 a better way of dealing with this, but for the moment (because 4.30 needs
4205 to be released), I have disabled address rewrite logging when running a
4206 filter in a non-root, non-exim process.
4212 1. The buildconfig auxiliary program wasn't quoting the value set for
4213 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4214 not defined. This bug was masked in 4.22 by the effect that was fixed in
4217 2. Some messages that were rejected after a message id was allocated were
4218 shown as "incomplete" by exigrep. It no longer does this for messages that
4219 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4221 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4222 have allow_domain_literals set, the ID did not get logged in the <= line.
4223 Domain literals are now always recognized in Message-ID: header lines.
4225 4. The first argument for a ${extract expansion item is the key name or field
4226 number. Leading and trailing spaces in this item were not being ignored,
4227 causing some misleading effects.
4229 5. When deliver_drop_privilege was set, single queue runner processes started
4230 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4231 same command in the process it spins off) were not dropping privilege.
4233 6. When the daemon running as "exim" started a queue runner, it always
4234 re-executed Exim in the spun-off process. This is a waste of effort when
4235 deliver_drop_privilege is set. The new process now just calls the
4236 queue-runner function directly.
4242 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4245 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4246 ignored. Though the use of -oP was forcing the writing of a pid file, it
4247 was always written to the default place.
4249 3. If the message "no IP address found for host xxxx" is generated during
4250 incoming verification, it is now followed by identification of the incoming
4251 connection (so you can more easily find what provoked it).
4253 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4255 5. Added some features to "harden" Exim a bit more against certain attacks:
4257 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4258 be put in Local/Makefile. This is like the never_users runtime option,
4259 but it cannot be overridden. The default setting is "root".
4261 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4262 prefix string with which any file named in a -C command line option
4265 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4266 is retained for -C and -D only if the caller of Exim is root. Without
4267 it, the exim user may also use -C and -D and retain privilege.
4269 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4270 command line option is disabled.
4272 6. Macro names set by the -D option must start with an upper case letter, just
4273 like macro names defined in the configuration file.
4275 7. Added "dereference=" facility to LDAP.
4277 8. Two instances of the typo "uknown" in the source files are fixed.
4279 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4280 the Configure-Makefile script screwed up while processing it.
4282 10. Incorporated PCRE 4.4.
4284 11. The SMTP synchronization check was not operating right at the start of an
4285 SMTP session. For example, it could not catch a HELO sent before the client
4286 waited for the greeting. There is now a check for outstanding input at the
4287 point when the greeting is written. Because of the duplex, asynchronous
4288 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4289 way, but not yet received, when the check is performed.
4291 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4292 on TCP/IP sockets, because this apparently causes some broken clients to
4295 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4296 unchanged) from the Cygwin maintainer.
4298 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4299 was supported instead of testing for mbx. Effectively a typo.
4301 15. The spa authenticator server code was not checking that the input it
4302 received was valid base64.
4304 16. The debug output line for the "set" modifier in ACLs was not showing the
4305 name of the variable that was being set.
4307 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4309 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4310 lookup on its own. Something else was needed to trigger the lookup. For
4311 example, a match in host_lookup or the need for a host name in a host list.
4312 Now, if $sender_host_name is referenced and the host name has not yet been
4313 looked up, a lookup is performed. If the lookup fails, the variable remains
4314 empty, and $host_lookup_failed is set to "1".
4316 19. Added "eqi" as a case-independent comparison operator.
4318 20. The saslauthd authentication condition could segfault if neither service
4319 nor realm was specified.
4321 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4322 error message that was logged was misleading, and incoming SMTP
4323 connections were dropped. The message is now more accurate, and temporary
4324 errors are given to SMTP connections.
4326 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4327 (except RSET) with a 421 error, until QUIT is received. However, it was
4328 failing to send a response to QUIT.
4330 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4331 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4332 ACL is now run after the helo_try_verify_hosts code.
4334 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4335 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4336 case-independent, so other case variants are also recognized). Apparently
4337 some systems use these upper case variants.
4339 25. If more than two messages were waiting for the same host, and a transport
4340 filter was specified for the transport, Exim sent two messages over the
4341 same TCP/IP connection, and then failed with "socket operation on non-
4342 socket" when it tried to send the third.
4344 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4346 27. The extern definition of crypt16() in expand.c was not being excluded when
4347 the OS had its own crypt16() function.
4349 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4350 as a preferred synonym for return_size_limit, both as an option and as an
4353 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4355 30. Changed the default configuration ACL to relax the local part checking rule
4356 for addresses that are not in any local domains. For these addresses,
4357 slashes and pipe symbols are allowed within local parts, but the sequence
4358 /../ is explicitly forbidden.
4360 31. SPA server authentication was not clearing the challenge buffer before
4363 32. log_message in a "warn" ACL statement was writing to the reject log as
4364 well as to the main log, which contradicts the documentation and doesn't
4365 seem right (because no rejection is happening). So I have stopped it.
4367 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4368 However, I am unable to do any testing of this.
4370 34. Fixed an infelicity in the appendfile transport. When checking directories
4371 for a mailbox, to see if any needed to be created, it was accidentally
4372 using path names with one or more superfluous leading slashes; tracing
4373 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4375 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4376 contents are added to the log line that is written for every discarded
4377 recipient. (Previously a log_message setting was ignored.)
4379 36. The ${quote: operator now quotes the string if it is empty.
4381 37. The install script runs exim in order to find its version number. If for
4382 some reason other than non-existence or emptiness, which it checks, it
4383 could not run './exim', it was installing it with an empty version number,
4384 i.e. as "exim-". This error state is now caught, and the installation is
4387 38. An argument was missing from the function that creates an error message
4388 when Exim fails to connect to the socket for saslauthd authentication.
4389 This could cause Exim to crash, or give a corrupted message.
4391 39. Added isip, isip4, and isip6 to ${if conditions.
4393 40. The ACL variables $acl_xx are now saved with the message, and can be
4394 accessed later in routers, transports, and filters.
4396 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4397 strings in the file are not string-expanded.
4399 42. If a MAIL command specified a SIZE value that was too large to fit into an
4400 int variable, the check against message_size_limit failed. Such values are
4401 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4402 day this will have to be increased, but I don't think I want to be around
4403 when emails are that large.
4410 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4411 iconv() is not standard in FreeBSD.
4413 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4414 IPv6 enabled. The observed symptom was a segmentation fault on return from
4415 the function os_common_find_running_interfaces() in src/os.c.
4417 3. In the check_special_case() function in daemon.c I had used "errno" as an
4418 argument name, which causes warnings on some systems. This was basically a
4419 typo, since it was named "eno" in the comments!
4421 4. The code that waits for the clock to tick (at a resolution of some fraction
4422 of a second) so as to ensure message-id uniqueness was always waiting for
4423 at least one whole tick, when it could have waited for less. [This is
4424 almost certainly not relevant at current processor speeds, where it is
4425 unlikely to ever wait at all. But we try to future-proof.]
4427 5. The function that sleeps for a time interval that includes fractions of a
4428 second contained a race. It did not block SIGALRM between setting the
4429 timer, and suspending (a couple of lines later). If the interval was short
4430 and the sigsuspend() was delayed until after it had expired, the suspension
4431 never ended. On busy systems this could lead to processes getting stuck for
4434 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4435 process, before it forks any delivery processes. The open lookup caching
4436 mechanism meant that the open file or database connection was passed into
4437 the delivery process. The problem was that delivery processes always tidy
4438 up cached lookup data. This could cause a problem for the next delivery
4439 process started by the queue runner, because the external queue runner
4440 process does not know about the closure. So the next delivery process
4441 still has data in the lookup cache. In the case of a file lookup, there was
4442 no problem because closing a file descriptor in a subprocess doesn't affect
4443 the parent. However, if the lookup was caching a connection to a database,
4444 the connection was closed, and the second delivery process was likely to
4445 see errors such as "PGSQL: query failed: server closed the connection
4446 unexpectedly". The problem has been fixed by closing all cached lookups
4447 in a queue runner before running a delivery process.
4449 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4450 seem to have the "const" qualifier which it has on other OS. I've
4453 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4454 *of the same type* (client or server) with the same public name that an
4455 error should be diagnosed.
4457 9. When Exim looked up a host name for an IP address, but failed to find the
4458 original IP address when looking up the host name (a safety check), it
4459 output the message "<ip address> does not match any IP for NULL", which was
4460 confusing, to say the least. The bug was that the host name should have
4461 appeared instead of "NULL".
4463 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4464 user that is built into the binary, and the -C or -D options is used, root
4465 privilege is dropped before the configuration file is read. In addition,
4466 logging is switched to stderr instead of the normal log files. If the
4467 configuration then re-defines the Exim user, the unprivileged environment
4468 is probably not what is expected, so Exim logs a panic warning message (but
4471 However, if deliver_drop_privilege is set, the unprivileged state may well
4472 be exactly what is intended, so the warning has been cut out in that case,
4473 and Exim is allowed to try to write to its normal log files.
4479 1. smtp_return_error_details was not giving details for temporary sender
4480 or receiver verification errors.
4482 2. Diagnose a configuration error if two authenticators have the same public
4485 3. Exim used not to create the message log file for a message until the first
4486 delivery attempt. This could be confusing when incoming messages were held
4487 for policy or load reasons. The message log file is now created at the time
4488 the message is received, and an initial "Received" line is written to it.
4490 4. The automatically generated man page for command line options had a minor
4491 bug that caused no ill effects; however, a more serious problem was that
4492 the procedure for building the man page automatically didn't always
4493 operate. Consequently, release 4.20 contains an out-of-date version. This
4494 shouldn't happen again.
4496 5. When building Exim with embedded Perl support, the script that builds the
4497 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4498 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4500 6. The freeze_tell option was not being used for messages that were frozen on
4501 arrival, either by an ACL or by local_scan().
4503 7. Added the smtp_incomplete_transaction log selector.
4505 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4506 was accepting AUTH without a new EHLO.
4508 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4509 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4512 10. Logging of TCP/IP connections (when configured) now happens in the main
4513 daemon process instead of the child process, so that the TCP/IP connection
4514 count is more accurate (but it can never be perfect).
4516 11. The use of "drop" in a nested ACL was not being handled correctly in the
4517 outer ACL. Now, if condition failure induced by the nested "drop" causes
4518 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4519 or "require"), the connection is dropped.
4521 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4522 that yield "discard" can now be used with an "accept" or a "discard" verb,
4523 but an error is generated for any others (because I can't see a useful way
4524 to define what should happen).
4526 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4527 are now processed in the same way as lines in the Exim configuration file.
4528 In particular, continuation lines are supported.
4530 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4532 15. Added -ti meaning -t -i.
4534 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4535 domains, and warn by logging if others are found.
4537 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4538 returned from GIFCONF when Exim is trying to find the list of interfaces on
4539 a host. The code in os.c has been modified to copy each ifreq to an aligned
4540 structure in all cases.
4542 Also, in some cases, the returned ifreq's were being copied to a 'struct
4543 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4544 means the last couple of bytes of an IPv6 address could be chopped if the
4545 ifreq contained only a normal sockaddr (14 bytes storage).
4547 18. Named domain lists were not supported in the hosts_treat_as_local option.
4548 An entry such as +xxxx was not recognized, and was treated as a literal
4551 19. Ensure that header lines added by a DATA ACL are included in the reject log
4552 if the ACL subsequently rejects the message.
4554 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4555 MD5 (which is deprecated).
4557 21. When testing a filter file using -bf, Exim was writing a message when it
4558 took the sender from a "From " line in the message, but it was not doing so
4559 when it took $return_path from a Return-Path: header line. It now does.
4561 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4562 with a valid header line field name (a series of printing characters
4563 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4565 23. Changed "disc" in the source to "disk" to conform to the documentation and
4566 the book and for uniformity.
4568 24. Ignore Sendmail's -Ooption=value command line item.
4570 25. When execve() failed while trying to run a command in a pipe transport,
4571 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4572 could be confused with a return value of 69 from the command itself. This
4573 has been changed to 127, the value the shell returns if it is asked to run
4574 a non-existent command. The wording for the related log line suggests a
4575 non-existent command as the problem.
4577 26. If received_header_text expands to an empty string, do not add a Received:
4578 header line to the message. (Well, it adds a token one on the spool, but
4579 marks it "old" so that it doesn't get used or transmitted.)
4581 27. Installed eximstats 1.28 (addition of -nt option).
4583 28. There was no check for failure on the call to getsockname() in the daemon
4584 code. This can fail if there is a shortage of resources on the system, with
4585 ENOMEM, for example. A temporary error is now given on failure.
4587 29. Contrary to the C standard, it seems that in some environments, the
4588 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4589 program. Exim now does this explicitly; it affects the formatting of
4590 timestamps using strftime().
4592 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4593 complaints. I've now initialized all the variables, to avoid this.
4595 32. Header lines added by a system filter were not being "seen" during
4596 transport-time rewrites.
4598 33. The info_callback() function passed to OpenSSL is set up with type void
4599 (*)(SSL *, int, int), as described somewhere. However, when calling the
4600 function (actually a macro) that sets it up, the type void(*)() is
4601 expected. I've put in a cast to prevent warnings from picky compilers.
4603 34. If a DNS black list lookup found a CNAME record, but there were no A
4604 records associated with the domain it pointed at, Exim crashed.
4606 35. If a DNS black list lookup returned more than one A record, Exim ignored
4607 all but the first. It now scans all returned addresses if a particular IP
4608 value is being sought. In this situation, the contents of the
4609 $dnslist_value variable are a list of all the addresses, separated by a
4612 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4613 to accept a host name and all its aliases if the forward lookup for any of
4614 them yielded the IP address of the incoming connection. Now it accepts only
4615 those names whose forward lookup yields the correct IP address. Any other
4616 names are discarded. This closes a loophole whereby a rogue DNS
4617 administrator could create reverse DNS records to break through a
4618 wildcarded host restriction in an ACL.
4620 37. If a user filter or a system filter that ran in a subprocess used any of
4621 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4622 the wrong values were passed to the pipe command ($thisaddress had the
4623 value of $0, $0 had the value of $1, etc). This bug was introduced by
4624 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4626 38. Improved the line breaking for long SMTP error messages from ACLs.
4627 Previously, if there was no break point between 40 and 75 characters, Exim
4628 left the rest of the message alone. Two changes have been made: (a) I've
4629 reduced the minimum length to 35 characters; (b) if it can't find a break
4630 point between 35 and 75 characters, it looks ahead and uses the first one
4631 that it finds. This may give the occasional overlong line, but at least the
4632 remaining text gets split now.
4634 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4635 file descriptors might be low, and that setting 1000 would always raise it.
4636 It turns out that in some environments, the limit is already over 1000 and
4637 that lowering it causes trouble. So now Exim takes care not to decrease it.
4639 40. When delivering a message, the value of $return_path is set to $sender_
4640 address at the start of routing (routers may change the value). By an
4641 oversight, this default was not being set up when an address was tested by
4642 -bt or -bv, which affected the outcome if any router or filter referred to
4645 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4646 log only when "message" or "log_message" are set. However, if one of the
4647 conditions was an address verification, or a call to a nested ACL, the
4648 messages generated by the underlying test were being passed through. This
4649 no longer happens. The underlying message is available in $acl_verify_
4650 message for both "message" and "log_message" expansions, so it can be
4651 passed through if needed.
4653 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4654 new expansion $bh_ to give the encoded byte string without charset
4655 translation. Translation happens only if iconv() is available; HAVE_ICONV
4656 indicates this at build time. HEADERS_CHARSET gives the charset to
4657 translate to; headers_charset can change it in the configuration, and
4658 "headers charset" can change it in an individual filter file.
4660 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4661 that creates RFC 2047 encoded "words" labels them as that charset instead
4662 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4663 expansion operator; (ii) when Exim creates a From: line for a local
4664 message; (iii) when a header line is rewritten to include a "phrase" part.
4666 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4667 buggy, causing it to skip the first lines of messages whose message ID
4668 ended in 'D'. This would not have bitten before Exim release 4.14, because
4669 message IDs were unlikely to end in 'D' before then. The effect was to have
4670 incorrect size information for certain domains.
4672 45. #include "config.h" was missing at the start of the crypt16.c module. This
4673 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4676 46. If there was a timeout during a "random" callout check, Exim treated it as
4677 a failure of the random address, and carried on sending RSET and the real
4678 address. If the delay was just some slowness somewhere, the response to the
4679 original RCPT would be taken as a response to RSET and so on, causing
4680 mayhem of various kinds.
4682 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4683 when I implemented it. It didn't allow for the fact that some option values
4684 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4685 the right test for positive values.
4687 48. Domain names in DNS records are case-independent. Exim always looks them up
4688 in lower case. Some resolvers return domain names in exactly the case they
4689 appear in the zone file, that is, they may contain uppercase letters. Not
4690 all resolvers do this - some return always lower case. Exim was treating a
4691 change of case by a resolver as a change of domain, similar to a widening
4692 of a domain abbreviation. This triggered its re-routing code and so it was
4693 trying to route what was effectively the same domain again. This normally
4694 caused routing to fail (because the router wouldn't handle the domain
4695 twice). Now Exim checks for this case specially, and just changes the
4696 casing of the domain that it ultimately uses when it transmits the message
4699 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4702 50. If a filter generated a file delivery with a non-absolute name (possible if
4703 no home directory exists for the router), the forbid_file option was not
4706 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4707 the existing equality matching.
4709 52. Exim was using ints instead of ino_t variables in some places where it was
4710 dealing with inode numbers.
4712 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4713 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4714 called TMPDIR, and if it finds it is different, it changes its value.
4716 54. The smtp_printf() function is now made available to local_scan() so
4717 additional output lines can be written before returning. There is also an
4718 smtp_fflush() function to enable the detection of a dropped connection.
4719 The variables smtp_input and smtp_batched_input are exported to
4722 55. Changed the default runtime configuration: the message "Unknown user"
4723 has been removed from the ACL, and instead placed on the localuser router,
4724 using the cannot_route_message feature. This means that any verification
4725 failures that generate their own messages won't get overridden. Similarly,
4726 the "Unrouteable address" message that was in the ACL for unverifiable
4727 relay addresses has also been removed.
4729 56. Added hosts_avoid_esmtp to the smtp transport.
4731 57. The exicyclog script was not checking for the esoteric option
4732 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4733 will work only if exicyclog is run under the appropriate euid.
4735 58. Following a discussion on the list, the rules by which Exim recognises line
4736 endings on incoming messages have been changed. The -dropcr and drop_cr
4737 options are now no-ops, retained only for backwards compatibility. The
4738 following line terminators are recognized: LF CRLF CR. However, special
4739 processing applies to CR:
4741 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4742 nor a local message in the state where . is a terminator.
4744 (ii) If a bare CR is encountered in a header line, an extra space is added
4745 after the line terminator so as not to end the header. The reasoning
4746 behind this is that bare CRs in header lines are most likely either
4747 to be mistakes, or people trying to play silly games.
4749 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4750 was being incorrectly given as 18 bytes larger than it should have been.
4751 This is a VOB (very old bug).
4753 60. This may never have affected anything current, but just in case it has:
4754 When the local host is found other than at the start of a list of hosts,
4755 the local host, those with the same MX, and any that follow, are discarded.
4756 When the list in question was part of a longer list of hosts, the following
4757 hosts (not currently being processed) were also being discarded. This no
4758 longer happens. I'm not sure if this situation could ever has previously
4761 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4764 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4765 header. This is something that is recommended in a new Internet Draft, and
4766 is something that is documented as being done by Sendmail. There are two
4767 possible values. For messages generated by the autoreply transport, Exim
4770 Auto-Submitted: auto-replied
4772 whereas for all other generated messages (e.g. bounces) it adds
4774 Auto-Submitted: auto-generated
4776 63. The "personal" condition in filters now includes a test for the
4777 Auto-Submitted: header. If it contains the string "auto-" the message it
4778 not considered personal.
4780 64. Added rcpt_include_affixes as a generic transport option.
4782 65. Added queue_only_override (default true).
4784 66. Added the syslog_duplication option.
4786 67. If what should have been the first header line of a message consisted of
4787 a space followed by a colon, Exim was mis-interpreting it as a header line.
4788 It isn't of course - it is syntactically invalid and should therefore be
4789 treated as the start of the message body. The misbehaviour could have
4790 caused a number of strange effects, including loss of data in subsequent
4791 header lines, and spool format errors.
4793 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4794 client host had authenticated. This control can now be exercised by an ACL
4795 for more flexibility.
4797 69. By default, callouts do not happen when testing with -bh. There is now a
4798 variant, -bhc, which does actually run the callout code, including
4799 consulting and updating the callout cache.
4801 70. Added support for saslauthd authentication, courtesy of Alexander
4804 71. If statvfs() failed on the spool or log directories while checking their
4805 size for availability, Exim confusingly gave the error "space shortage".
4806 Furthermore, in debugging mode it crashed with a floating point exception.
4807 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4808 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4809 serious problem, Exim now writes to the main and panic logs when this
4810 happens, with details of the failure. It then refuses to accept the
4811 incoming message, giving the message "spool directory problem" or "log
4812 directory problem" with a 421 code for SMTP messages.
4814 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4815 0, 1, and 2 exist, because some OS complain for execs without them (see
4816 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4817 descriptors. However, the code omitted to check that the open succeeded,
4818 causing mysterious errors if for some reason the permissions on /dev/null
4819 got screwed. Now Exim writes a message to the main and panic logs, and
4820 bombs out if it can't open /dev/null.
4822 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4823 interact so that it is all more flexible. It is supposed to remain
4824 backwards compatible. Also added extra_local_interfaces.
4826 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4827 to bomb out with an assertion failure - to the client this appears as a
4828 connection drop. This problem occurs in the part of the code that was taken
4829 from the Samba project. Fortunately, the assertion is in a very simple
4830 function, so I have fixed this by reproducing the function inline in the
4831 one place where it is called, and arranging for authentication to fail
4832 instead of killing the process with assert().
4834 75. The SPA client code was not working when the server requested OEM rather
4835 than Unicode encoding.
4837 76. Added code to make require_files with a specific uid setting more usable in
4838 the case where statting the file as root fails - usually a non-root-mounted
4839 NFS file system. When this happens and the failure is EACCES, Exim now
4840 forks a subprocess and does the per-uid checking as the relevant uid.
4842 77. Added process_log_path.
4844 78. If log_file_path was not explicitly set, a setting of check_log_space or
4845 check_log_inodes was ignored.
4847 79. If a space check for the spool or log partitions fails, the incident is now
4848 logged. Of course, in the latter case the data may get lost...
4850 80. Added the %p formatting code to string_format() so that it can be used to
4851 print addresses in debug_print(). Adjusted all the address printing in the
4852 debugging in store.c to use %p rather than %d.
4854 81. There was a concern that a line of code in smtp_in.c could overflow a
4855 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4856 initially expressed, the concern was not well-founded, because trailing
4857 spaces are removed early. However, if the trailing spaces were followed by
4858 a NULL, they did not get removed, so the overflow was possible. Two fixes
4861 (a) I re-wrote the offending code in a cleaner fashion.
4862 (b) If an incoming SMTP command contains a NULL character, it is rejected
4865 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4866 runs initgroups(), so that if the Exim user is in any additional groups,
4867 they will be used during message reception.
4873 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob