1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.534 2007/10/18 12:00:43 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
6 TK/01 Add preliminary DKIM support. Currently requires a forked version of
7 ALT-N's libdkim that I have put here:
8 http://duncanthrax.net/exim-experimental/
10 Note to Michael Haardt: I had to rename some vars in sieve.c. They
11 were called 'true' and it seems that C99 defines that as a reserved
12 keyword to be used with 'bool' variable types. That means you could
13 not include C99-style headers which use bools without triggering
14 build errors in sieve.c.
16 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
17 as mailq or other aliases. Changed the --help handling significantly
18 to do whats expected. exim_usage() emits usage/help information.
20 SC/01 Added the -bylocaldomain option to eximstats.
22 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
27 PH/01 Another patch from the Sieve maintainer.
29 PH/02 When an IPv6 address is converted to a string for single-key lookup
30 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
31 dots are used instead of colons so that keys in lsearch files need not
32 contain colons. This was done some time before quoting was made available
33 in lsearch files. However, iplsearch files do require colons in IPv6 keys
34 (notated using the quote facility) so as to distinguish them from IPv4
35 keys. This meant that lookups for IP addresses in host lists did not work
36 for iplsearch lookups.
38 This has been fixed by arranging for IPv6 addresses to be expressed with
39 colons if the lookup type is iplsearch. This is not incompatible, because
40 previously such lookups could never work.
42 The situation is now rather anomolous, since one *can* have colons in
43 ordinary lsearch keys. However, making the change in all cases is
44 incompatible and would probably break a number of configurations.
46 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
49 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
50 conversion specification without a maximum field width, thereby enabling
51 a rogue spamd server to cause a buffer overflow. While nobody in their
52 right mind would setup Exim to query an untrusted spamd server, an
53 attacker that gains access to a server running spamd could potentially
54 exploit this vulnerability to run arbitrary code as the Exim user.
56 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
57 $primary_hostname instead of what libspf2 thinks the hosts name is.
59 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
60 a directory entry by the name of the lookup key. Previously, if a
61 symlink pointed to a non-existing file or a file in a directory that
62 Exim lacked permissions to read, a lookup for a key matching that
63 symlink would fail. Now it is enough that a matching directory entry
64 exists, symlink or not. (Bugzilla 503.)
66 PH/03 The body_linecount and body_zerocount variables are now exported in the
69 PH/04 Added the $dnslist_matched variable.
71 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
72 This means they are set thereafter only if the connection becomes
75 PH/06 Added the client_condition to authenticators so that some can be skipped
76 by clients under certain conditions.
78 PH/07 The error message for a badly-placed control=no_multiline_responses left
79 "_responses" off the end of the name.
81 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
83 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
84 (without spaces) instead of just copying the configuration text.
86 PH/10 Added the /noupdate option to the ratelimit ACL condition.
88 PH/11 Added $max_received_linelength.
90 PH/12 Added +ignore_defer and +include_defer to host lists.
92 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
93 way in which PCRE > 7.0 is built.
95 PH/14 Implemented queue_only_load_latch.
97 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
98 MAIL command. The effect was to mangle the value on 64-bit systems.
100 PH/16 Another patch from the Sieve maintainer.
102 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
104 PH/18 If a system quota error occurred while trying to create the file for
105 a maildir delivery, the message "Mailbox is full" was not appended to the
106 bounce if the delivery eventually timed out. Change 4.67/27 below applied
107 only to a quota excession during the actual writing of the file.
109 PH/19 It seems that peer DN values may contain newlines (and other non-printing
110 characters?) which causes problems in log lines. The DN values are now
111 passed through string_printing() before being added to log lines.
113 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
114 and InterBase are left for another time.)
116 PH/21 Added message_body_newlines option.
118 PH/22 Guard against possible overflow in moan_check_errorcopy().
120 PH/23 POSIX allows open() to be a macro; guard against that.
122 PH/24 If the recipient of an error message contained an @ in the local part
123 (suitably quoted, of course), incorrect values were put in $domain and
124 $local_part during the evaluation of errors_copy.
130 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
131 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
134 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
135 issue a MAIL command.
137 PH/02 In an ACL statement such as
139 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
141 if a client was not listed at all, or was listed with a value other than
142 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
143 the condition was not true (as it should be), so access was not denied.
144 The bug was that the ! inversion was incorrectly passed on to the second
145 item. This has been fixed.
147 PH/03 Added additional dnslists conditions == and =& which are different from
148 = and & when the dns lookup returns more than one IP address.
150 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
151 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
153 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
154 FSYNC, which compiles an option called disable_fsync that allows for
155 bypassing fsync(). The documentation is heavily laced with warnings.
157 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
159 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
160 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
161 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
162 including adding "make clean"; (3) Added -fPIC when compiling the test
163 dynamically loaded module, to get rid of a warning.
165 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
166 message fails, move_frozen_messages = true and ignore_bounce_errors_after
167 = 0s. The bug is otherwise harmless.
169 PH/07 There was a bug in the dovecot authenticator such that the value of
170 $auth1 could be overwritten, and so not correctly preserved, after a
171 successful authentication. This usually meant that the value preserved by
172 the server_setid option was incorrect.
174 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
176 PH/09 Installed PCRE release 7.0.
178 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
179 run for batched SMTP input. It is now run at the start of every message
180 in the batch. While fixing this I discovered that the process information
181 (output by running exiwhat) was not always getting set for -bs and -bS
182 input. This is fixed, and it now also says "batched" for BSMTP.
184 PH/11 Added control=no_pipelining.
186 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
187 patch, slightly modified), and move the expansion of helo_data till after
188 the connection is made in the smtp transport (so it can use these
191 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
193 PH/14 Added log_selector = +pid.
195 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
197 PH/16 Add ${if forany and ${if forall.
199 PH/17 Added dsn_from option to vary the From: line in DSNs.
201 PH/18 Flush SMTP output before performing a callout, unless control =
202 no_callout_flush is set.
204 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
205 was true (the default) a successful delivery failed to delete the retry
206 item, thus causing premature timeout of the address. The bug is now
209 PH/20 Added hosts_avoid_pipelining to the smtp transport.
211 PH/21 Long custom messages for fakedefer and fakereject are now split up
212 into multiline reponses in the same way that messages for "deny" and
213 other ACL rejections are.
215 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
216 with slight modification.
218 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
219 draft, changing the syntax and factoring some duplicate code".
221 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
222 for deliveries of the second and subsequent messages over the same SMTP
225 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
226 ${reduce, with only minor "tidies".
228 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
230 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
231 expansion side effects.
233 PH/27 When a message times out after an over-quota error from an Exim-imposed
234 quota, the bounce message says "mailbox is full". This message was not
235 being given when it was a system quota that was exceeded. It now should
238 MH/03 Made $recipients available in local_scan(). local_scan() already has
239 better access to the recipient list through recipients_list[], but
240 $recipients can be useful in postmaster-provided expansion strings.
242 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
243 in the case of a MAIL command with additional options following the
244 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
245 were accidentally chopped off.
247 PH/29 SMTP synchronization checks are implemented when a command is read -
248 there is a check that no more input is waiting when there shouldn't be
249 any. However, for some commands, a delay in an ACL can mean that it is
250 some time before the response is written. In this time, more input might
251 arrive, invalidly. So now there are extra checks after an ACL has run for
252 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
253 pipelining has not been advertised.
255 PH/30 MH's patch to allow iscntrl() characters to be list separators.
257 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
258 returned in the SMTP response when smtp_return_error_details was false.
261 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
262 instead of the C I/O that was originally supplied, because problems were
265 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
266 Exim which did not show up earlier: it was assuming that a call to
267 SSL_CTX_set_info_callback() might give an error value. In fact, there is
268 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
269 was a macro that became an assignment, so it seemed to work. This has
270 changed to a proper function call with a void return, hence the compile
271 error. Exim's code has been fixed.
273 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
276 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
278 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
280 PH/37 If a message is not accepted after it has had an id assigned (e.g.
281 because it turns out to be too big or there is a timeout) there is no
282 "Completed" line in the log. When some messages of this type were
283 selected by exigrep, they were listed as "not completed". Others were
284 picked up by some special patterns. I have improved the selection
285 criteria to be more general.
287 PH/38 The host_find_failed option in the manualroute router can now be set
288 to "ignore", to completely ignore a host whose IP address cannot be
289 found. If all hosts are ignored, the behaviour is controlled by the new
290 host_all_ignored option.
292 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
293 homing or because of multiple MX records with /mx) generated more than
294 one IP address, and the following item turned out to be the local host,
295 all the secondary addresses of the first item were incorrectly removed
296 from the list, along with the local host and any following hosts (which
297 is what is supposed to happen).
299 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
300 whoever called Exim into the -H file. In the case of the daemon it was
301 behaving confusingly. When first started, it used values for whoever
302 started the daemon, but after a SIGHUP it used the Exim user (because it
303 calls itself on a restart). I have changed the code so that it now always
306 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
307 message are rejected with the same error (e.g. no authentication or bad
308 sender address), and a DATA command is nevertheless sent (as can happen
309 with PIPELINING or a stupid MUA), the error message that was given to the
310 RCPT commands is included in the rejection of the DATA command. This is
311 intended to be helpful for MUAs that show only the final error to their
314 PH/42 Another patch from the Sieve maintainer.
316 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
317 Eximstats - Fixed some broken HTML links and added missing column headers
319 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
320 columns for Rejects, Temp Rejects, Ham, and Spam rows.
322 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
324 PH/43 Yet another patch from the Sieve maintainer.
326 PH/44 I found a way to check for a TCP/IP connection going away before sending
327 the response to the final '.' that terminates a message, but only in the
328 case where the client has not sent further data following the '.'
329 (unfortunately, this is allowed). However, in many cases there won't be
330 any further data because there won't be any more messages to send. A call
331 to select() can be used: if it shows that the input is "ready", there is
332 either input waiting, or the socket has been closed. An attempt to read
333 the next input character can distinguish the two cases. Previously, Exim
334 would have sent an OK response which the client would never have see.
335 This could lead to message repetition. This fix should cure that, at
336 least in a lot of common cases.
338 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
339 advertised in response to EHLO.
345 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
346 fixed by 4.65/MH/01 (is this a record?) are fixed:
348 (i) An empty string was always treated as zero by the numeric comparison
349 operators. This behaviour has been restored.
351 (ii) It is documented that the numeric comparison operators always treat
352 their arguments as decimal numbers. This was broken in that numbers
353 starting with 0 were being interpreted as octal.
355 While fixing these problems I realized that there was another issue that
356 hadn't been noticed. Values of message_size_limit (both the global option
357 and the transport option) were treated as octal if they started with 0.
358 The documentation was vague. These values are now always treated as
359 decimal, and I will make that clear in the documentation.
365 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
366 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
369 MH/01 Don't check that the operands of numeric comparison operators are
370 integers when their expansion is in "skipping" mode (fixes bug
371 introduced by 4.64-PH/07).
373 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
374 child addresses, Exim now panics and dies. Previously, because the count
375 is held in a short int, deliveries were likely to be lost. As such a
376 large number of recipients for a single message is ridiculous
377 (performance will be very, very poor), I have chosen to impose a limit
378 rather than extend the field.
384 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
385 leftover -K file (the existence of which was triggered by #402).
386 While we were at it, introduced process PID as part of the -K
387 filename. This should rule out race conditions when creating
390 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
391 processing considerably. Previous code took too long for large mails,
392 triggering a timeout which in turn triggers #401.
394 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
395 in the DK code in transports.c. sendfile() is not really portable,
396 hence the _LINUX specificness.
398 TF/01 In the add_headers option to the mail command in an Exim filter,
399 there was a bug that Exim would claim a syntax error in any
400 header after the first one which had an odd number of characters
403 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
404 callout verification, Exim cached a "reject" for the entire domain. This
405 is correct for most verifications, but it is not correct for a recipient
406 verification with use_sender or use_postmaster set, because in that case
407 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
408 case of MAIL FROM:<> rejection from other early rejections (e.g.
409 rejection of HELO). When verifying a recipient using a non-null MAIL
410 address, the cache is ignored if it shows MAIL FROM:<> rejection.
411 Whatever the result of the callout, the value of the domain cache is
412 left unchanged (for any other kind of callout, getting as far as trying
413 RCPT means that the domain itself is ok).
415 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
418 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
419 manifest itself as EPIPE rather than ECONNECT. When tidying away a
420 session, the daemon ignores ECONNECT errors and logs others; it now
421 ignores EPIPE as well.
423 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
424 (quoted-printable decoding).
426 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
427 later the small subsequent patch to fix an introduced bug.
429 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
431 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
433 PH/08 An error is now given if message_size_limit is specified negative.
435 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
436 to be given (somewhat) arbitrary names.
438 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
441 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
442 miscellaneous code fixes
444 PH/10 Added the log_reject_target ACL modifier to specify where to log
447 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
448 hostname. This is wrong, because it relates to the incoming message (and
449 probably the interface on which it is arriving) and not to the outgoing
450 callout (which could be using a different interface). This has been
451 changed to use the value of the helo_data option from the smtp transport
452 instead - this is what is used when a message is actually being sent. If
453 there is no remote transport (possible with a router that sets up host
454 addresses), $smtp_active_hostname is used.
456 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
457 tweaks were necessary in order to get it to work (see also 21 below):
458 (a) The code assumed that strncpy() returns a negative number on buffer
459 overflow, which isn't the case. Replaced with Exim's string_format()
461 (b) There were several signed/unsigned issues. I just did the minimum
462 hacking in of casts. There is scope for a larger refactoring.
463 (c) The code used strcasecmp() which is not a standard C function.
464 Replaced with Exim's strcmpic() function.
465 (d) The code set only $1; it now sets $auth1 as well.
466 (e) A simple test gave the error "authentication client didn't specify
467 service in request". It would seem that Dovecot has changed its
468 interface. Fortunately there's a specification; I followed it and
469 changed what the client sends and it appears to be working now.
471 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
474 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
475 address A is aliased to B and C, where B exists and C does not. Without
476 -v the output is "A verified" because verification stops after a
477 successful redirection if more than one address is generated. However,
478 with -v the child addresses are also verified. Exim was outputting "A
479 failed to verify" and then showing the successful verification for C,
480 with its parentage. It now outputs "B failed to verify", showing B's
481 parentage before showing the successful verification of C.
483 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
484 look up a TXT record in a specific list after matching in a combined
487 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
488 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
489 they consult the DNS. I had assumed they would set it the way they
490 wanted; and indeed my experiments on Linux seem to show that in some
491 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
492 To be on the safe side, however, I have now made the interface to
493 host_find_byname() similar to host_find_bydns(), with an argument
494 containing the DNS resolver options. The host_find_byname() function now
495 sets these options at its start, just as host_find_bydns() does. The smtp
496 transport options dns_qualify_single and dns_search_parents are passed to
497 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
498 of host_find_byname() use the default settings of RES_DEFNAMES
499 (qualify_single) but not RES_DNSRCH (search_parents).
501 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
502 spool_read_header() do less string testing, by means of a preliminary
503 switch on the second character of optional "-foo" lines. (This is
504 overdue, caused by the large number of possibilities that now exist.
505 Originally there were few.) While I was there, I also converted the
506 str(n)cmp tests so they don't re-test the leading "-" and the first
507 character, in the hope this might squeeze out yet more improvement.
509 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
510 flag allowing group syntax was set by the header_syntax check but not
511 turned off, possible causing trouble later; (2) The flag was not being
512 set at all for the header_verify test, causing "group"-style headers to
513 be rejected. I have now set it in this case, and also caused header_
514 verify to ignore an empty address taken from a group. While doing this, I
515 came across some other cases where the code for allowing group syntax
516 while scanning a header line wasn't quite right (mostly, not resetting
517 the flag correctly in the right place). These bugs could have caused
518 trouble for malformed header lines. I hope it is now all correct.
520 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
521 with the "reply" argument non-NULL. The code, however (which originally
522 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
523 but it didn't always do it. This confused somebody who was copying the
524 code for some other use. I have removed all the tests.
526 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
527 feature that was used to support insecure browsers during the U.S. crypto
528 embargo. It requires special client support, and Exim is probably the
529 only MTA that supported it -- and would never use it because real RSA is
530 always available. This code has been removed, because it had the bad
531 effect of slowing Exim down by computing (never used) parameters for the
532 RSA_EXPORT functionality.
534 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
535 authenticator to fail if there's a tab character in the incoming data
536 (there should never be unless someone is messing about, as it's supposed
537 to be base64-encoded). Also added, on Timo's advice, the "secured" option
538 if the connection is using TLS or if the remote IP is the same as the
539 local IP, and the "valid-client-cert option" if a client certificate has
542 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
543 authenticators. This can be used for authorization after authentication
544 succeeds. (In the case of plaintext, it servers for both authentication
547 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
548 if any retry times were supplied.
550 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
551 connection, where there is no client IP address to check. In this
552 situation, the verify now always succeeds.
554 PH/25 Applied John Jetmore's -Mset patch.
556 PH/26 Added -bem to be like -Mset, but loading a message from a file.
558 PH/27 In a string expansion for a processed (not raw) header when multiple
559 headers of the same name were present, leading whitespace was being
560 removed from all of them, but trailing whitespace was being removed only
561 from the last one. Now trailing whitespace is removed from each header
562 before concatenation. Completely empty headers in a concatenation (as
565 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
566 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
568 PH/29 [Removed. This was a change that I later backed out, and forgot to
569 correct the ChangeLog entry (that I had efficiently created) before
570 committing the later change.]
572 PH/30 Exim was sometimes attempting to deliver messages that had suffered
573 address errors (4xx response to RCPT) over the same connection as other
574 messages routed to the same hosts. Such deliveries are always "forced",
575 so retry times are not inspected. This resulted in far too many retries
576 for the affected addresses. The effect occurred only when there were more
577 hosts than the hosts_max_try setting in the smtp transport when it had
578 the 4xx errors. Those hosts that it had tried were not added to the list
579 of hosts for which the message was waiting, so if all were tried, there
580 was no problem. Two fixes have been applied:
582 (i) If there are any address or message errors in an SMTP delivery, none
583 of the hosts (tried or untried) are now added to the list of hosts
584 for which the message is waiting, so the message should not be a
585 candidate for sending over the same connection that was used for a
586 successful delivery of some other message. This seems entirely
587 reasonable: after all the message is NOT "waiting for some host".
588 This is so "obvious" that I'm not sure why it wasn't done
589 previously. Hope I haven't missed anything, but it can't do any
590 harm, as the worst effect is to miss an optimization.
592 (ii) If, despite (i), such a delivery is accidentally attempted, the
593 routing retry time is respected, so at least it doesn't keep
594 hammering the server.
596 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
597 in ${readsocket because some servers need this prod.
599 PH/32 Added some extra debug output when updating a wait-xxx database.
601 PH/33 The hint "could be header name not terminated by colon", which has been
602 given for certain expansion errors for a long time, was not being given
603 for the ${if def:h_colon_omitted{... case.
605 PH/34 The spec says: "With one important exception, whenever a domain list is
606 being scanned, $domain contains the subject domain." There was at least
607 one case where this was not true.
609 PH/35 The error "getsockname() failed: connection reset by peer" was being
610 written to the panic log as well as the main log, but it isn't really
611 panic-worthy as it just means the connection died rather early on. I have
612 removed the panic log writing for the ECONNRESET error when getsockname()
615 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
616 runs only) independently of the message's sender address. This meant
617 that, if the 4xx error was in fact related to the sender, a different
618 message to the same recipient with a different sender could confuse
619 things. In particualar, this can happen when sending to a greylisting
620 server, but other circumstances could also provoke similar problems.
621 I have changed the default so that the retry time for these errors is now
622 based a combination of the sender and recipient addresses. This change
623 can be overridden by setting address_retry_include_sender=false in the
626 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
627 remote server are returned as part of bounce messages. This was not
628 happening for LMTP over a pipe (the lmtp transport), but now it is the
629 same for both kinds of LMTP.
631 PH/38 Despite being documented as not happening, Exim was rewriting addresses
632 in header lines that were in fact CNAMEs. This is no longer the case.
634 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
635 and queue runs started by the daemon processed all messages. This has
636 been fixed so that -R and -S can now usefully be given with -q<time>.
638 PH/40 Import PCRE release 6.7 (fixes some bugs).
640 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
642 PH/42 Give an error if -q is specified more than once.
644 PH/43 Renamed the variables $interface_address and $interface_port as
645 $received_ip_address and $received_port, to make it clear that these
646 values apply to message reception, and not to the outgoing interface when
647 a message is delivered. (The old names remain recognized, of course.)
649 PH/44 There was no timeout on the connect() call when using a Unix domain
650 socket in the ${readsocket expansion. There now is.
652 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
653 be meaningful with "accept".
655 SC/01 Eximstats V1.43
656 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
658 SC/02 Eximstats V1.44
659 Use a glob alias rather than an array ref in the generated
660 parser. This improves both readability and performance.
662 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
663 Collect SpamAssassin and rejection statistics.
664 Don't display local sender or destination tables unless
665 there is data to show.
666 Added average volumes into the top table text output.
668 SC/04 Eximstats V1.46
669 Collect data on the number of addresses (recipients)
670 as well as the number of messages.
672 SC/05 Eximstats V1.47
673 Added 'Message too big' to the list of mail rejection
674 reasons (thanks to Marco Gaiarin).
676 SC/06 Eximstats V1.48
677 Mainlog lines which have GMT offsets and are too short to
678 have a flag are now skipped.
680 SC/07 Eximstats V1.49 (Alain Williams)
681 Added the -emptyok flag.
683 SC/08 Eximstats V1.50
684 Fixes for obtaining the IP address from reject messages.
686 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
687 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
688 whitesspace changes from 4.64-PH/27
690 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
693 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
694 are found, allow negative numbers in numeric criteria)
696 JJ/06 exipick.20061117.2, added new $message_body_missing variable
698 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
699 to match changes made in 4.64-PH/43
701 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
703 PH/47 Put in an explicit test for a DNS lookup of an address record where the
704 "domain" is actually an IP address, and force a failure. This locks out
705 those revolvers/nameservers that support "A-for-A" lookups, in
706 contravention of the specifications.
708 PH/48 When a host name was looked up from an IP address, and the subsequent
709 forward lookup of the name timed out, the host name was left in
710 $sender_host_name, contrary to the specification.
712 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
713 restricted to single-key lookups, Exim was not diagnosing an error if
714 * or *@ was used with a query-style lookup.
716 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
718 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
719 long ago, but noone interested enough thought of it. Let's just say that
720 the "1.1" means that there are some new functions that weren't there at
721 some point in the past.
723 PH/51 Error processing for expansion failure of helo_data from an smtp
724 transport during callout processing was broken.
726 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
727 tested/used via the -bh/-bhc/-bs options.
729 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
730 bug, fixed in subsequent PCRE releases).
732 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
733 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
735 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
741 SC/01 Use a glob alias rather than an array ref in eximstats generated
742 parser. This improves both readability and performance.
744 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
745 Don't display local sender or destination tables in eximstats unless
746 there is data to show.
747 Added average volumes into the eximstats top table text output.
749 SC/03 Collect data on the number of addresses (recipients) as well
750 as the number of messages in eximstats.
752 TF/01 Correct an error in the documentation for the redirect router. Exim
753 does (usually) call initgroups() when daemonizing.
755 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
756 with consistent privilege compared to when running as a daemon.
758 TF/03 Note in the spec that $authenticated_id is not set for local
759 submissions from trusted users.
761 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
762 Thanks to Dean Brooks <dean@iglou.com> for the patch.
764 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
765 by adding some example configuration directives to the default
766 configuration file. A little bit of work is required to uncomment the
767 directives and define how usernames and passwords are checked, but
768 there is now a framework to start from.
770 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
771 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
772 without this. I don't know how relevant this is to other LDAP libraries.
774 PH/02 Add the verb name to the "unknown ACL verb" error.
776 PH/03 Magnus Holmgren's patch for filter_prepend_home.
778 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
780 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
781 directory not expanded when it should be if an expanded home directory
782 was set for the address (which is overridden by the transport).
784 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
787 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
788 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
789 because it is too late at that time, and has no effect.
791 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
792 security issue with \' (bugzilla #107). I could not use the
793 PQescapeStringConn() function, because it needs a PGconn value as one of
796 PH/08 When testing addresses using -bt, indicate those final addresses that
797 are duplicates that would not cause an additional delivery. At least one
798 person was confused, thinking that -bt output corresponded to deliveries.
799 (Suppressing duplicates isn't a good idea as you lose the information
800 about possibly different redirections that led to the duplicates.)
802 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
803 systems where poll() doesn't work, in particular OS X.
805 PH/10 Added more information to debugging output for retry time not reached.
807 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
808 operations in malware.c.
810 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
813 PH/13 If write_rejectlog was set false when logging was sent to syslog with
814 syslog_duplication set false, log lines that would normally be written
815 both the the main log and to the reject log were not written to syslog at
818 PH/14 In the default configuration, change the use of "message" in ACL warn
819 statements to "add_header".
821 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
822 not followed by a command (e.g. "seen endif").
824 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
825 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
828 PH/17 Added extra conditions to the default value of delay_warning_condition
832 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
833 { match{$h_precedence:}{(?i)bulk|list|junk} } \
834 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
837 The Auto-Submitted: and various List- headers are standardised, whereas I
838 don't think Precedence: ever was.
840 PH/18 Refactored debugging code in route_finduser() to show more information,
841 in particular, the error code if getpwnam() issues one.
843 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
844 This is apparently needed in addition to the PH/07 change above to avoid
845 any possible encoding problems.
847 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
848 but not after initializing Perl.
850 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
851 output them only if debugging. By default they are written stderr,
852 apparently, which is not desirable.
854 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
857 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
860 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
862 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
863 authenticated or an ident call has been made. Suppress the default
864 values for $authenticated_id and $authenticated_sender (but permit -oMai
865 and -oMas) when testing with -bh.
867 PH/24 Re-jigged the order of the tests in the default configuration so that the
868 tests for valid domains and recipients precede the DNS black list and CSA
869 tests, on the grounds that those ones are more expensive.
871 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
872 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
873 This bug exists in every version of Exim that I still have, right back to
876 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
877 However, an attempt to turn on case-sensitivity in a regex key by
878 including (?-i) didn't work because the subject string was already
879 lowercased, and the effects were non-intuitive. It turns out that a
880 one-line patch can be used to allow (?-i) to work as expected.
886 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
887 other effects) broke the use of negated acl sub-conditions.
889 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
892 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
893 "Deny" causes Exim to reject the incoming connection with a 554 error.
894 Unfortunately, if there is a major crisis, such as a disk failure,
895 tcp-wrappers gives "deny", whereas what one would like would be some
896 kind of temporary error. A kludge has been added to help with this.
897 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
898 554 error is used if errno is still zero or contains ENOENT (which occurs
899 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
902 PH/03 Add -lutil to the default FreeBSD LIBS setting.
904 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
905 errors. Otherwise a message that provokes a temporary error (when other
906 messages do not) can cause a whole host to time out.
908 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
909 addresses were routed directly to files or pipes from a redirect router.
910 File deliveries just didn't batch; pipe deliveries might have suffered
913 PH/06 A failure to get a lock for a hints database would erroneously always say
914 "Failed to get write lock", even when it was really a read lock.
916 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
917 of 0600. This has been changed to use the value of the lockfile_mode
918 option (which defaults to 0600).
920 PH/08 Applied small patch from the Sieve maintainer.
922 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
923 folder from quota calculations, a direct delivery into this folder messed
924 up the contents of the maildirsize file. This was because the regex was
925 used only to exclude .Trash (or whatever) when the size of the mailbox
926 was calculated. There was no check that a delivery was happening into an
927 excluded directory. This bug has been fixed by ignoring all quota
928 processing for deliveries into excluded directories.
930 PH/10 Added the maildirfolder_create_regex option to appendfile.
936 PH/01 The code for finding all the local interface addresses on a FreeBSD
937 system running IPv6 was broken. This may well have applied to all BSD
938 systems, as well as to others that have similar system calls. The broken
939 code found IPv4 interfaces correctly, but gave incorrect values for the
940 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
941 that it would not match correctly against @[] and not recognize the IPv6
944 PH/02 The ipliteral router was not recognizing addresses of the form user@
945 [ipv6:....] because it didn't know about the "ipv6:" prefix.
947 PH/03 Added disable_ipv6.
949 PH/04 Changed $reply_address to use the raw form of the headers instead of the
950 decoded form, because it is most often used to construct To: headers
951 lines in autoreplies, and the decoded form may well be syntactically
952 invalid. However, $reply_address has leading white space removed, and all
953 newlines turned into spaces so that the autoreply transport does not
956 PH/05 If group was specified without a user on a router, and no group or user
957 was specified on a transport, the group from the router was ignored.
959 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
960 for visible compile-time settings that can be used to change these
961 numbers, for those that want even more. Backwards compatibility with old
962 spool files has been maintained. However, going back to a previous Exim
963 release will lost any variables that are in spool files.
965 PH/07 Two small changes when running in the test harness: increase delay when
966 passing a TCP/IP connection to a new process, in case the original
967 process has to generate a bounce, and remove special handling of
968 127.0.0.2 (sic), which is no longer necessary.
970 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
971 be the same on different OS.
973 PH/09 Moved a debug statement in filter processing to avoid a race problem when
976 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
977 whether --show-vars was specified or not
979 JJ/02 exipick: Added support for new ACL variable spool format introduced
982 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
983 syntactically invalid From: or Reply-to: line, and a filter used this to
984 generate an autoreply, and therefore failed to obtain an address for the
985 autoreply, Exim could try to deliver to a non-existent relative file
986 name, causing unrelated and misleading errors. What now happens is that
987 it logs this as a hard delivery error, but does not attempt to create a
990 PH/11 The exinext utility has a -C option for testing purposes, but although
991 the given file was scanned by exinext itself; it wasn't being passed on
992 when Exim was called.
994 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
995 an end-of-file indication when reading a command response.
997 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
998 compiled. In many other places in Exim, IPv6 addresses are always
999 recognized, so I have changed this. It also means that IPv4 domain
1000 literals of the form [IPV4:n.n.n.n] are now always recognized.
1002 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1003 used if the router is not running as root, for example, when verifying at
1004 ACL time, or when using -bh. The debugging output from this situation was
1005 non-existent - all you got was a failure to exec. I have made two
1008 (a) Failures to set uid/gid, the current directory, or a process leader
1009 in a subprocess such as that created by queryprogram now generate
1010 suitable debugging ouput when -d is set.
1012 (b) The queryprogram router detects when it is not running as root,
1013 outputs suitable debugging information if -d is set, and then runs
1014 the subprocess without attempting to change uid/gid.
1016 PH/15 Minor change to Makefile for building test_host (undocumented testing
1019 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1020 additional section of a DNS packet that returns MX or SRV records.
1021 Instead, it always explicitly searches for A/AAAA records. This avoids
1022 major problems that occur when a DNS server includes only records of one
1023 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1024 fixed another bug: if SRV records were looked up and the corresponding
1025 address records were *not* found in the additional section, the port
1026 values from the SRV records were lost.
1028 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1029 using the correct key (the original address) when searching the retry
1030 rules in order to find which one to use for generating the retry hint.
1032 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1033 adding the default one. Similarly, if it contains a Reply-To: header, the
1034 errors_reply_to option, if set, is not used.
1036 PH/19 When calculating a retry time, Exim used to measure the "time since
1037 failure" by looking at the "first failed" field in the retry record. Now
1038 it does not use this if it is later than than the arrival time of the
1039 message. Instead it uses the arrival time. This makes for better
1040 behaviour in cases where some deliveries succeed, thus re-setting the
1041 "first failed" field. An example is a quota failure for a huge message
1042 when small messages continue to be delivered. Without this change, the
1043 "time since failure" will always be short, possible causing more frequent
1044 delivery attempts for the huge message than are intended.
1045 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1047 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1048 $1, $2, $3) because the numerical variables can be reset during some
1049 expansion items (e.g. "match"), thereby losing the authentication data.
1051 PH/21 Make -bV show the size of off_t variables so that the test suite can
1052 decide whether to run tests for quotas > 2G.
1054 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1055 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1056 filecount value is greater than 2G or if a quota value is greater than 2G
1057 on a system where the size of off_t is not greater than 4, a panic error
1060 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1061 never match. The debug and -bh output now contains an explicit error
1062 message indicating a malformed IPv4 address or mask.
1064 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1065 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1066 PH/23 above applies.
1068 PH/25 Do not write to syslog when running in the test harness. The only
1069 occasion when this arises is a failure to open the main or panic logs
1070 (for which there is an explicit test).
1072 PH/26 Added the /no_tell option to "control=freeze".
1074 PH/27 If a host name lookup failed very early in a connection, for example, if
1075 the IP address matched host_lookup and the reverse lookup yielded a name
1076 that did not have a forward lookup, an error message of the form "no IP
1077 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1078 could be logged. Now it outputs the IP address instead of "NULL".
1080 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1081 allows the sender and the authenticated sender to be set when
1082 submitting a message from within Exim. Since child_open_exim() is
1083 documented for local_scan(), the new function should be too.
1085 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1086 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1087 results in an empty string is now treated as unset.
1089 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1091 PH/31 Added sender_verify_fail logging option.
1093 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1094 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1095 tidied the source and removed it altogether.
1097 PH/33 When a queue run was abandoned because the load average was too high, a
1098 log line was always written; now it is written only if the queue_run log
1099 selector is set. In addition, the log line for abandonment now contains
1100 information about the queue run such as the pid. This is always present
1101 in "start" and "stop" lines but was omitted from the "abandon" line.
1103 PH/34 Omit spaces between a header name and the colon in the error message that
1104 is given when verify = headers_syntax fails (if there are lots of them,
1105 the message gets confusing).
1107 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1108 names, as there are now some PTR records that contain slashes. This check
1109 is only to protect against broken name servers that fall over on strange
1110 characters, so the fact that it applies to all lookups doesn't matter.
1112 PH/36 Now that the new test suite is complete, we can remove some of the
1113 special code in Exim that was needed for the old test suite. For example,
1114 sorting DNS records because real resolvers return them in an arbitrary
1115 order. The new test suite's fake resolver always returns records in the
1118 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1119 bounces) except when queue_only is set, to avoid logging races between
1120 the different processes.
1122 PH/38 Panic-die if .include specifies a non-absolute path.
1124 PH/39 A tweak to the "H" retry rule from its user.
1126 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1127 a label. They prevented compilation on older perls.
1129 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1130 a warning to be raised on newish perls.
1132 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1133 on queue. Changes to match documented behaviour of showing count of
1134 messages matching specified criteria.
1136 PH/40 Changed the default ident timeout from 30s to 5s.
1138 PH/41 Added support for the use of login_cap features, on those BSD systems
1139 that have them, for controlling the resources used by pipe deliveries.
1141 PH/42 The content-scanning code uses fopen() to create files in which to put
1142 message data. Previously it was not paying any attention to the mode of
1143 the files. Exim runs with umask(0) because the rest of the code creates
1144 files with open(), and sets the required mode explicitly. Thus, these
1145 files were ending up world-writeable. This was not a big issue, because,
1146 being within the spool directory, they were not world-accessible. I have
1147 created a function called modefopen, which takes an additional mode
1148 argument. It sets umask(777), creates the file, chmods it to the required
1149 mode, then resets the umask. All the relevant calls to fopen() in the
1150 content scanning code have been changed to use this function.
1152 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1153 to 24 hours. This avoids potential overflow problems when processing G
1154 and H retry rules. I suspect nobody ever tinkers with this value.
1156 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1158 PH/45 When the plaintext authenticator is running as a client, the server's
1159 challenges are checked to ensure they are valid base64 strings. By
1160 default, the authentication attempt is cancelled if an invalid string is
1161 received. Setting client_ignore_invalid_base64 true ignores these errors.
1162 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1163 they are received. Thus, the responses can be made to depend on the
1164 challenges. If an invalid string is ignored, an empty string is placed in
1167 PH/46 Messages that are created by the autoreply transport now contains a
1168 References: header, in accordance with RFCs 2822 and 3834.
1170 PH/47 Added authenticated_sender_force to the smtp transport.
1172 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1174 PH/49 Installed latest patch from the Sieve maintainer.
1176 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1177 was also set, the appendfile transport was unnecessarily scanning a
1178 directory of message files (e.g. for maildir delivery) to find the count
1179 of files (along with the size), even though it did not need this
1180 information. It now does the scan only if it needs to find either the
1181 size of the count of files.
1183 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1185 PH/52 Two bugs concerned with error handling when the smtp transport is
1188 (i) Exim was not creating retry information for temporary errors given
1189 for individual recipients after the DATA command when the smtp transport
1190 was used in LMTP mode. This meant that they could be retried too
1191 frequently, and not timed out correctly.
1193 (ii) Exim was setting the flag that allows error details to be returned
1194 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1195 recipients that were returned after the DATA command.
1197 PH/53 This is related to PH/52, but is more general: for any failing address,
1198 when detailed error information was permitted to be returned to the
1199 sender, but the error was temporary, then after the final timeout, only
1200 "retry timeout exceeded" was returned. Now it returns the full error as
1201 well as "retry timeout exceeded".
1203 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1204 do this, and (what is worse) MTAs that accept it.
1206 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1207 will now be deprecated.
1209 PH/56 New os.c-cygwin from the Cygwin maintainer.
1211 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1212 formats (previously only available in exim formats via -bpr, -bpru,
1213 and -bpra. Now also available in native and exiqgrep formats)
1215 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1216 with very large, slow to parse queues
1218 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1220 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1222 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1223 responses to authentication challenges, though it was showing the
1224 challenges; (ii) I've removed the CR characters from the debug output for
1227 PH/58 Allow for the insertion of a newline as well as a space when a string
1228 is turned into more than one encoded-word during RFC 2047 encoding. The
1229 Sieve code now uses this.
1231 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1232 data_4xx, lost_connection, tls_required.
1234 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1235 message was being sent as an SMTP response.
1237 PH/61 Add -l and -k options to exicyclog.
1239 PH/62 When verifying, if an address was redirected to one new address, so that
1240 verification continued, and the new address failed or deferred after
1241 having set something in $address_data, the value of $address_data was not
1242 passed back to the ACL. This was different to the case when no
1243 redirection occurred. The value is now passed back in both cases.
1245 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1246 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1247 use login_cap.h, so on its own it isn't the distinguishing feature. The
1248 new name refers directly to the setclassresources() function.
1250 PH/65 Added configuration files for NetBSD3.
1252 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1254 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1255 is preferred over IPv4.
1257 PH/68 The bounce_return_message and bounce_return_body options were not being
1258 honoured for bounces generated during the reception of non-SMTP messages.
1259 In particular, this applied to messages rejected by the ACL. This bug has
1260 been fixed. However, if bounce_return_message is true and bounce_return_
1261 body is false, the headers that are returned for a non-SMTP message
1262 include only those that have been read before the error was detected.
1263 (In the case of an ACL rejection, they have all been read.)
1265 PH/69 The HTML version of the specification is now built in a directory called
1266 spec_html instead of spec.html, because the latter looks like a path with
1267 a MIME-type, and this confuses some software.
1269 PH/70 Catch two compiler warnings in sieve.c.
1271 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1272 function verify_get_ident() calls ip_connect() to connect a socket, but
1273 if the "connect()" function timed out, ip_connect() used to close the
1274 socket. However, verify_get_ident() also closes the socket later, and in
1275 between Exim writes to the log, which may get opened at this point. When
1276 the socket was closed in ip_connect(), the log could get the same file
1277 descriptor number as the socket. This naturally causes chaos. The fix is
1278 not to close the socket in ip_connect(); the socket should be closed by
1279 the function that creates it. There was only one place in the code where
1280 this was missing, in the iplookup router, which I don't think anybody now
1281 uses, but I've fixed it anyway.
1283 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1284 well as to direct DNS lookups. Otherwise the handling of names in host
1285 lists is inconsistent and therefore confusing.
1291 PH/01 Two changes to the default runtime configuration:
1293 (1) Move the checks for relay_from_hosts and authenticated clients from
1294 after to before the (commented out) DNS black list checks.
1296 (2) Add control=submission to the relay_from_hosts and authenticated
1297 clients checks, on the grounds that messages accepted by these
1298 statements are most likely to be submissions.
1300 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1302 (1) Generate an error if the third argument for the ${prvs expansion is
1305 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1308 (3) Reset the variables that are obtained from the first argument of
1309 ${prvscheck and used in the second argument before leaving the code,
1310 because their memory is reclaimed, so using them afterwards may do
1313 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1314 one (it's much easier than Tom thought :-).
1316 (5) Because of (4), we can now allow for the use of $prvscheck_result
1317 inside the third argument.
1319 PH/03 For some reason, the default setting of PATH when running a command from
1320 a pipe transport was just "/usr/bin". I have changed it to
1323 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1324 anything to be listed in the output from -bV.
1326 PH/05 When a filter generated an autoreply, the entire To: header line was
1327 quoted in the delivery log line, like this:
1329 => >A.N.Other <ano@some.domain> <original@ddress> ...
1331 This has been changed so that it extracts the operative address. There
1332 may be more than one such address. If so, they are comma-separated, like
1335 => >ano@some.domain,ona@other.domain <original@ddress> ...
1337 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1338 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1339 not being looked up in the rDNS to get a host name, Exim was showing the
1340 IP address twice in Received: lines, even though the IP addresses were
1341 identical. For example:
1343 Received: from [1.2.3.4] (helo=[1.2.3.4])
1345 However, if the real host name was known, it was omitting the HELO data
1346 if it matched the actual IP address. This has been tidied up so that it
1347 doesn't show the same IP address twice.
1349 PH/07 When both +timestamp and +memory debugging was on, the value given by
1350 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1351 called by the debug printing, thereby overwriting the timestamp buffer.
1352 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1354 PH/08 When the original message was included in an autoreply transport, it
1355 always said "this is a copy of the message, including all the headers",
1356 even if body_only or headers_only was set. It now gives an appropriate
1359 PH/09 Applied a patch from the Sieve maintainer which:
1361 o fixes some comments
1362 o adds the (disabled) notify extension core
1363 o adds some debug output for the result of if/elsif tests
1364 o points to the current vacation draft in the documentation
1365 and documents the missing references header update
1369 o fixes a bug in processing the envelope test (when testing
1370 multiple envelope elements, the last element determinted the
1373 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1374 Electronic Mail") by including:
1376 Auto-submitted: auto-generated
1378 in the messages that it generates (bounce messages and others, such as
1379 warnings). In the case of bounce messages for non-SMTP mesages, there was
1380 also a typo: it was using "Auto_submitted" (underscore instead of
1381 hyphen). Since every message generated by Exim is necessarily in response
1382 to another message, thes have all been changed to:
1384 Auto-Submitted: auto-replied
1386 in accordance with these statements in the RFC:
1388 The auto-replied keyword:
1390 - SHOULD be used on messages sent in direct response to another
1391 message by an automatic process,
1393 - MUST NOT be used on manually-generated messages,
1395 - MAY be used on Delivery Status Notifications (DSNs) and Message
1396 Disposition Notifications (MDNs),
1398 - MUST NOT be used on messages generated by automatic or periodic
1399 processes, except for messages which are automatic responses to
1402 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1403 to the default Received: header definition.
1405 PH/12 Added log selector acl_warn_skipped (default on).
1407 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1408 variables because (a) they are in the wrong storage pool and (b) even if
1409 they were copied, it wouldn't work properly because of the caching.
1411 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1412 checking when decoding. Apparently there are clients that generate
1413 overlong encoded strings. Why am I not surprised?
1415 PH/15 If the first argument of "${if match_address" was not empty, but did not
1416 contain an "@" character, Exim crashed. Now it writes a panic log message
1417 and treats the condition as false.
1419 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1421 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1422 extension "envelope-auth". The code is finished and in agreement with
1423 other implementations, but there is no documentation so far and in fact,
1424 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1425 not changing the active code.
1427 Print executed "if" and "elsif" statements when debugging is used. This
1428 helps a great deal to understand what a filter does.
1430 Document more things not specified clearly in RFC3028. I had all this
1431 sorted out, when out of a sudden new issues came to my mind. Oops."
1433 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1436 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1437 canonical form (as documented). However, after a host name lookup from
1438 the IP address, check_host() was doing a simple string comparison with
1439 addresses acquired from the DNS when checking that the found name did
1440 have the original IP as one of its addresses. Since any found IPv6
1441 addresses are likely to be in abbreviated form, the comparison could
1442 fail. Luckily, there already exists a function for doing the comparison
1443 by converting both addresses to binary, so now that is used instead of
1444 the text comparison.
1446 PH/20 There was another similar case to PH/19, when a complete host name was
1447 given in a host list; looking up its IP address could give an abbreviated
1448 form, whereas the current host's name might or might not be abbreviated.
1449 The same fix has been applied.
1455 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1456 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1459 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1460 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1462 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1464 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1465 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1466 Also, for "Auto-Submitted", treat anything other than "no" as
1467 non-personal, in accordance with RFC 3834. (Previously it treated
1468 anything starting "auto-" as non-personal.)
1470 TF/01 The control=submission/name=... option had a problem with syntax
1471 errors if the name included a slash character. The /name= option
1472 now slurps the rest of the string, so it can include any characters
1473 but it must come last in the list of options (after /sender_retain
1476 PH/05 Some modifications to the interface to the fake nameserver for the new
1484 TK/01 Added the "success_on_redirect" address verification option. See
1485 NewStuff for rationale and an example.
1487 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1489 PH/02 Patch to exigrep to allow it to work on syslog lines.
1491 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1492 fread() to skip over the body file's header line, because in Cygwin the
1493 header line is locked and is inaccessible.
1495 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1496 co-exist for some time) to make it clear that it is the Exim ID that is
1497 referenced, not the Message-ID: header line.
1499 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1500 string_format() function, because snprintf() does not exist on all
1503 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1504 ${stat: expansion item.
1506 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1507 protocol synchronization error", to keep the pedants happy.
1509 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1510 well as for IRIX systems, when gcc is being used. See the host.c source
1513 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1515 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1518 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1519 transport and to the smtp transport in LMTP mode.
1521 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1523 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1524 run a filter in a subprocess. This could lead to confusion in subsequent
1525 lookups in the parent process. There should also be a search_tidyup() at
1526 the end of the subprocess.
1528 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1529 only if the host matched helo_try_verify_hosts, which caused the
1530 verification to occur when the EHLO/HELO command was issued. The ACL just
1531 tested the remembered result. Now, if a previous verification attempt has
1532 not happened, "verify = helo" does it there and then.
1534 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1536 TK/03 Fix log output including CR from clamd.
1538 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1539 exist provoked a memory error which could cause a segfault.
1541 PH/15 Installed PCRE 6.2
1543 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1545 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1546 of the problem. Specifically, suggested +O2 rather than +O1 for the
1549 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1551 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1552 record, the port was not correctly logged when the outgoing_port log
1553 selector was set (it logged the transort's default port).
1555 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1556 fallback_hosts, and "hosts" in the smtp transport.
1558 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1559 host errors such as "Connection refused".
1561 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1562 authentication with radiusclient 0.4.9:
1564 - Error returned from rc_read_config was caught wrongly
1565 - Username/password not passed on to radius server due to wrong length.
1567 The presumption is that some radiusclient API changes for 4.51/PH/17
1568 were not taken care of correctly. The code is still untested by me (my
1569 Linux distribution still has 0.3.2 of radiusclient), but it was
1570 contributed by a Radius user.
1572 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1573 expanding the "port" option of the smtp transport.
1575 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1576 while reading a MIME header. Thanks to Tom Hughes for a patch.
1578 PH/24 Include config.h inside local_scan.h so that configuration settings are
1581 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1582 that in an ACL for RCPT (for example), you can examine exactly what was
1585 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1586 commands, but it was not correctly comparing the address with the actual
1587 client host address. Thus, it would show the EHLO address in Received:
1588 header lines when this was not necessary.
1590 PH/27 Added the % operator to ${eval:}.
1592 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1593 it should be ignoring failures (because with -C, for example, it has lost
1594 privilege). It wasn't ignoring creation failures other than "already
1597 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1600 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1601 because an input line was too long, either on its own, or by virtue of
1602 too many continuations, the temporary file was not being removed, and the
1603 return code was incorrect.
1605 PH/31 Missing "BOOL" in function definition in filtertest.c.
1607 PH/32 Applied Sieve patches from the maintainer.
1609 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1611 PH/33 Added "verify = not_blind".
1613 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1614 Local/Makefile (with some defaults set). These are used in built scripts
1615 such as exicyclog, but they have never been used in the exim_install
1616 script (though there are many overriding facilities there). I have
1617 arranged that the exim_install script now takes note of these two
1620 PH/35 Installed configuration files for Dragonfly.
1622 PH/36 When a locally submitted message by a trusted user did not contain a
1623 From: header, and the sender address was obtained from -f or from an SMTP
1624 MAIL command, and the trusted user did not use -F to supply a sender
1625 name, $originator_name was incorrectly used when constructing a From:
1626 header. Furthermore, $originator_name was used for submission mode
1627 messages from external hosts without From: headers in a similar way,
1628 which is clearly wrong.
1630 PH/37 Added control=suppress_local_fixups.
1632 PH/38 When log_selector = +received_sender was set, and the addition of the
1633 sender made the log line's construction buffer exactly full, or one byte
1634 less than full, an overflow happened when the terminating "\n" was
1637 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1638 when the result of a list match is failure because a DNS lookup failed.
1640 PH/40 RM_COMMAND is now used in the building process.
1642 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1643 the "build-* directories that it finds.
1645 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1646 address in a domain literal was a prefix of an interface address.
1648 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1649 when verifying a sender address, unless rewrite_headers is false.
1651 PH/44 Wrote a long comment about why errors_to addresses are verified as
1652 recipients, not senders.
1654 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1655 the ratelimit ACL was added.
1657 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1659 PH/46 Added extra information about PostgreSQL errors to the error string.
1661 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1662 avoiding the need to install special zones in a real server. This is
1663 backwards compatible; if it can't find the fake resolver, it drops back.
1664 Thus, both old and new test suites can be run.
1666 TF/02 Added util/ratelimit.pl
1668 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1669 clock is set back in time.
1671 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1672 Candler <B.Candler@pobox.com>.
1674 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1675 OK for addresses that are the result of redirections.
1677 PH/48 A number of further additions for the benefit of the new test suite,
1678 including a fake gethostbyname() that interfaces to the fake DNS resolver
1681 TF/06 The fix for widen_domains has also been applied to qualify_single and
1682 search_parents which are the other dnslookup options that can cause
1685 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1688 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1690 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1691 service" if it encountered an MX record with an empty target hostname.
1692 The message is now "an MX or SRV record indicated no SMTP service".
1694 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1695 if the DNS of the sending site is misconfigured. This is quite a
1696 common situation. This change restores the behaviour of treating a
1697 helo verification defer as a failure.
1699 PH/51 If self=fail was set on a router, the bounce message did not include the
1700 actual error message.
1706 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1708 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1709 command itself ended in error, the underlying message about the transport
1710 filter timeout was being overwritten with the pipe command error. Now the
1711 underlying error message should be appended to the second error message.
1713 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1715 PH/02 Reduce the amount of output that "make" produces by default. Full output
1716 can still be requested.
1718 PH/03 The warning log line about a condition test deferring for a "warn" verb
1719 was being output only once per connection, rather than after each
1720 occurrence (because it was using the same function as for successful
1721 "warn" verbs). This seems wrong, so I have changed it.
1723 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1724 it should not have, which might have caused a crash in the right
1725 circumstances, but probably never did.
1727 PH/04 Installed a modified version of Tony Finch's patch to make submission
1728 mode fix the return path as well as the Sender: header line, and to
1729 add a /name= option so that you can make the user's friendly name appear
1732 TF/03 Added the control = fakedefer ACL modifier.
1734 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1735 Mark Lowes for thorough testing.
1737 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1739 TK/03 Merged latest SRS patch from Miles Wilton.
1741 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1742 with the definition in sysexits.h (which is #included earlier).
1743 Fortunately, Exim does not actually use EX_OK. The code used to try to
1744 preserve the sysexits.h value, by assumimg that macro definitions were
1745 scanned for macro replacements. I have been disabused of this notion,
1746 so now the code just undefines EX_OK before #including unistd.h.
1748 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1749 in the smtp transport. When a block could not be written in a single
1750 write() function, the timeout was being re-applied to each part-write.
1751 This seems wrong - if the receiver was accepting one byte at a time it
1752 would take for ever. The timeout is now adjusted when this happens. It
1753 doesn't have to be particularly precise.
1755 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1756 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1758 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1759 without a domain if the check to <postmaster@domain> fails.
1761 SC/01 Eximstats: added -xls and the ability to specify output files
1762 (patch written by Frank Heydlauf).
1764 SC/02 Eximstats: use FileHandles for outputing results.
1766 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1768 SC/04 Eximstats: fixed display of large numbers with -nvr option
1770 SC/05 Eximstats: fixed merging of reports with empty tables.
1772 SC/06 Eximstats: added the -include_original_destination flag
1774 SC/07 Eximstats: removed tabs and trailing whitespace.
1776 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1778 TK/06 MBOX spool code: Add real "From " MBOX separator line
1779 so the .eml file is really in mbox format (even though
1780 most programs do not really care). Patch from Alex Miller.
1782 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1783 The latter is generated from $received_to and is only set if the
1784 message has one envelope recipient. SA can use these headers,
1785 obviously out-of-the-box. Patch from Alex Miller.
1787 PH/08 The ${def test on a variable was returning false if the variable's
1788 value was "0", contrary to what the specification has always said!
1789 The result should be true unless the variable is empty.
1791 PH/09 The syntax error of a character other than { following "${if
1792 def:variable_name" (after optional whitespace) was not being diagnosed.
1793 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1794 accidental colon was present, for example, could give incorrect results.
1796 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1797 result is used (not including appendfile, where other changes are about
1800 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1801 This involved changing a lot of size variables from int to off_t. It
1802 should work with maildirs and everything.
1804 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1805 spamd dying while we are connected to it.
1807 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1810 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1811 function for BDB 4.3.
1813 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1815 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1816 the sources for PCRE have been re-arranged and more files are now
1819 PH/15 The code I had for printing potentially long long variables in PH/11
1820 above was not the best (it lost precision). The length of off_t variables
1821 is now inspected at build time, and an appropriate printing format (%ld
1822 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1823 to be "long long int" or "long int". This is needed for the internal
1824 formatting function string_vformat().
1826 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1827 the configuration file to be ":syslog", then the script "guesses" where
1828 the logs files are, rather than using the compiled in default. In our
1829 case the guess is not the same as the compiled default, so the script
1830 suddenly stopped working when I started to use syslog. The patch checks
1831 to see if log_file_path is "". If so, it attempts to read it from exim
1832 with no configuration file to get the compiled in version, before it
1833 falls back to the previous guessing code."
1835 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1836 implementing BATV in an Exim configuration. See NewStuff for the gory
1839 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1840 Makefile that are specific to HP-UX.
1842 PH/18 If the "use_postmaster" option was set for a recipient callout together
1843 with the "random" option, the postmaster address was used as the MAIL
1844 FROM address for the random test, but not for the subsequent recipient
1845 test. It is now used for both.
1847 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1848 patch removes a few documentation additions to RFC 3028, because the
1849 latest draft now contains them. It adds the new en;ascii-case comparator
1850 and a new error check for 8bit text in MIME parts. Comparator and
1851 require names are now matched exactly. I enabled the subaddress
1852 extension, but it is not well tested yet (read: it works for me)."
1854 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1855 rework some of the code of TK/09 above to avoid the hardwired use of
1856 "%lld" and "long long". Replaced the call to snprintf() with a call to
1859 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1860 records point to non-existent hosts", "retry timeout exceeded", and
1861 "retry time not reached for any host after a long failure period".
1863 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1864 experimental DomainKeys support:
1866 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1867 (2) On an error such as an illegally used "control", the wrong name for
1868 the control was given.
1870 These problems did NOT occur unless DomainKeys support was compiled.
1872 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1874 PH/24 Added ${if match_ip condition.
1876 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1877 databases so that it will be absolutely obvious if a crash occurs in the
1878 DB library. This is a regular occurrence (often caused by mis-matched
1881 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1882 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1883 picked up on a user's system that detects such things. There doesn't seem
1884 to be a gcc warning option for this - only an attribute that has to be
1885 put on the function's prototype. It seems that in Fedora Core 4 they have
1886 set this on a number of new functions. No doubt there will be more in due
1889 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1890 specify a transport. However, if an address that was verified by such a
1891 router was the subject of a callout, Exim crashed because it tried to
1892 read the rcpt_include_affixes from the non-existent transport. Now it
1893 just assumes that the setting of that option is false. This bug was
1894 introduced by 4.51/PH/31.
1896 PH/28 Changed -d+all to exclude +memory, because that information is very
1897 rarely of interest, but it makes the output a lot bigger. People tend to
1898 do -d+all out of habit.
1900 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1901 code in os-type was giving problems when libc.so lives in lib64, like on
1904 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1905 aren't the modern standard, and it seems that some systems' include files
1906 don't always have them. Exim was already checking for some of the newer
1907 ones like T_AAAA, and defining it itself. I've added checks for all the
1908 record types that Exim uses.
1910 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1911 not automatically generating a new one, as it is supposed to. This
1912 prevented TLS from working. If the file did exist, but contained invalid
1913 data, a new version was generated, as expected. It was only the case of a
1914 non-existent file that was broken.
1916 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1917 with a change in libdomainkeys > 0.64.
1919 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1920 from DNS. If the selector record carries the flag, it now has
1921 precedence over the domain-wide flag.
1923 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1925 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1926 the use of a port name that isn't defined in /etc/services) occurred, the
1927 message was deferred as in a normal delivery, and thus remained on the
1928 spool, instead of being failed because of the mua_wrapper setting. This
1929 is now fixed, and I tidied up some of the mua_wrapper messages at the
1932 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1933 the messages in a hash of arrays rather than using individual hashes.
1934 This is a bit cleaner and results in dramatic memory savings, albeit
1935 at a slight CPU cost.
1937 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1938 as requested by Marc Sherman.
1940 SC/10 Eximstats: added histograms for user specified patterns as requested
1943 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1945 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1946 fopen() in the content-scanning modules that did not already have it.
1952 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1953 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1955 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1957 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1959 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1960 to test Sieve filters that use "vacation".
1962 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1963 that changes the way the GnuTLS parameters are stored in the cache file.
1964 The new format can be generated externally. For backward compatibility,
1965 if the data in the cache doesn't make sense, Exim assumes it has read an
1966 old-format file, and it generates new data and writes a new file. This
1967 means that you can't go back to an older release without removing the
1970 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1971 work if there are any delivery delays because "one_time" forces the
1972 parent to be marked "delivered", so its unseen clone is never tried
1973 again. For this reason, Exim now forbids the simultaneous setting of
1976 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1977 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1978 entry if you want to know the details. The fix, however, neglected to
1979 consider the case when local delivery batching is involved. The test for
1980 "previously delivered" was not happening when checking to see if an
1981 address could be batched with a previous (undelivered) one; under
1982 certain circumstances this could lead to multiple deliveries to the same
1985 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1986 in its include files, and this causes problems building Exim.
1988 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1989 header_syntax) but Exim was just ignoring anything given after a slash.
1990 In particular, this caused confusion with an attempt to use "verify =
1991 reverse_host_lookup/defer_ok". An error is now given when options are
1992 supplied for verify items that do not have them. (Maybe reverse_host_
1993 lookup should have a defer_ok option, but that's a different point.)
1995 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1996 defined by RFC 821) to 2048, because there were problems with some AUTH
1997 commands, and RFC 1869 says the size should be increased for extended
1998 SMTP commands that take arguments.
2000 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2003 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2004 "unknown" error; now it says that the functionality isn't in the binary.
2006 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2007 an address' error message when a string expansion fails (syntax or
2008 whatever). Otherwise the password may appear in the log. Following change
2009 PH/42 below, there is no longer a chance of it appearing in a bounce
2012 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2014 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2015 of its addresses was ever tried. (Bugzilla bug #2.)
2017 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2018 the result incorrectly in the debug output. (It correctly added a newline
2019 to what was transported.)
2021 TF/01 Added $received_time.
2023 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2024 commented out examples of how to interface to a virus scanner and to
2025 SpamAssassin. Also added commented examples of av_scanner and
2026 spamd_address settings.
2028 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2029 and controls are allowed in which ACLs. There were a couple of minor
2030 errors. Some of the entries in the conditions table (which is a table of
2031 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2032 negation of where the condition IS allowed.
2034 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2036 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2037 header file does not have a version number, so I've had to invent a new
2038 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2039 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2040 radiusclient), but it was contributed by a Radius user.
2042 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2043 files or directories, for OpenSSL.
2045 PH/19 When an Exim process that is running as root has to create an Exim log
2046 file, it does so in a subprocess that runs as exim:exim so as to get the
2047 ownership right at creation (otherwise, other Exim processes might see
2048 the file with the wrong ownership). There was no test for failure of this
2049 fork() call, which would lead to the process getting stuck as it waited
2050 for a non-existent subprocess. Forks do occasionally fail when resources
2051 run out. I reviewed all the other calls to fork(); they all seem to check
2054 PH/20 When checking for unexpected SMTP input at connect time (before writing
2055 the banner), Exim was not dealing correctly with a non-positive return
2056 from the read() function. If the client had disconnected by this time,
2057 the result was a log entry for a synchronization error with an empty
2058 string after "input=" when read() returned zero. If read() returned -1
2059 (an event I could not check), uninitialized data bytes were printed.
2060 There were reports of junk text (parts of files, etc) appearing after
2063 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2065 PH/22 Added support for macro redefinition, and (re)definition in between
2066 driver and ACL definitions.
2068 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2069 forgetting to use the resulting value; it was using the unexpanded value.
2071 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2072 hadn't been configured. The fix is from Juergen Kreileder, who
2073 understands it better than I do:
2075 "Here's what I see happening with three configured cyrus_sasl
2076 authenticators configured (plain, login, cram-md5):
2078 On startup auth_cyrus_sasl_init() gets called for each of these.
2079 This means three calls to sasl_listmech() without a specified mech_list.
2080 => SASL tests which mechs of all available mechs actually work
2081 => three warnings about OTP not working
2082 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2084 With the patch, sasl_listmech() also gets called three times. But now
2085 SASL's mech_list option is set to the server_mech specified in the the
2086 authenticator. Or in other words, the answer from sasl_listmech()
2087 gets limited to just the mech you're testing for (which is different
2089 => the return list contains just 'plain' or 'login', 'cram-md5' or
2090 nothing depending on the value of ob->server_mech.
2092 I've just tested the patch: Authentication still works fine,
2093 unavailable mechs specified in the exim configuration are still
2094 caught, and the auth.log warnings about OTP are gone."
2096 PH/25 When debugging is enabled, the contents of the command line are added
2097 to the debugging output, even when log_selector=+arguments is not
2100 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2101 answer is "GNU", and only if the return is "GNU/something" is the answer
2104 PH/27 $acl_verify_message is now set immediately after the failure of a
2105 verification in an ACL, and so is available in subsequent modifiers. In
2106 particular, the message can be preserved by coding like this:
2108 warn !verify = sender
2109 set acl_m0 = $acl_verify_message
2111 Previously, $acl_verify_message was set only while expanding "message"
2112 and "log_message" when a very denied access.
2114 PH/28 Modified OS/os.c-Linux with
2116 -#ifndef OS_LOAD_AVERAGE
2117 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2119 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2120 nomenclature these days.)
2122 PH/29 Installed patch from the Sieve maintainer that adds the options
2123 sieve_useraddress and sieve_subaddress to the redirect router.
2125 PH/30 In these circumstances:
2126 . Two addresses routed to the same list of hosts;
2127 . First host does not offer TLS;
2128 . First host accepts first address;
2129 . First host gives temporary error to second address;
2130 . Second host offers TLS and a TLS session is established;
2131 . Second host accepts second address.
2132 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2133 and peerdn, if requested) that were in fact used only for the second
2136 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2137 attention to any local part prefix or suffix that was matched by the
2138 router that accepted the address. It now behaves in the same way as it
2139 does for delivery: the affixes are removed from the local part unless
2140 rcpt_include_affixes is set on the transport.
2142 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2143 timeout during the DATA phase of an incoming message.
2145 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2146 applied a patch sanctioned by the Sieve maintainer.
2148 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2149 the uid or gid is negative. A case of a negative gid caused this to be
2150 noticed. The fix allows for either to be negative.
2152 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2153 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2156 PH/36 The change PH/12 above was broken. Fixed it.
2158 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2159 the grounds that routing the same address twice would always produce the
2160 same answer. This might have been true once, but it is certainly no
2161 longer true now. Routing a child address may depend on the previous
2162 routing that produced that child. Some complicated redirection strategies
2163 went wrong when messages had multiple recipients, and made Exim's
2164 behaviour dependent on the order in which the addresses were given.
2166 I have moved the duplicate checking until after the routing is complete.
2167 Exim scans the addresses that are assigned to local and remote
2168 transports, and removes any duplicates. This means that more work will be
2169 done, as duplicates will always all be routed, but duplicates are
2170 presumably rare, so I don't expect this is of any significance.
2172 For deliveries to pipes, files, and autoreplies, the duplicate checking
2173 still happens during the routing process, since they are not going to be
2176 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2177 It corrects a timeout issue with spamd. This is Ian's comment: "The
2178 background is that sometimes spamd either never reads data from a
2179 connection it has accepted, or it never writes response data. The exiscan
2180 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2181 blindly assumes that writes won't block so it may never time out."
2183 PH/39 Allow G after quota size as well as K and M.
2185 PH/40 The value set for $authenticated_id in an authenticator may not contain
2186 binary zeroes or newlines because the value is written to log lines and
2187 to spool files. There was no check on this. Now the value is run through
2188 the string_printing() function so that such characters are converted to
2189 printable escape sequences.
2191 PH/41 $message_linecount is a new variable that contains the total number of
2192 lines in the message. Compare $body_linecount, which is the count for the
2195 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2196 bounce and delay warning messages, except in certain special cases, which
2199 (a) An SMTP error message from a remote host;
2200 (b) A message specified in a :fail: redirection;
2201 (c) A message specified in a "fail" command in a system filter;
2202 (d) A message specified in a FAIL return from the queryprogram router;
2203 (e) A message specified by the cannot_route_message router option.
2205 In these cases only, Exim does include the error details in bounce and
2206 warning messages. There are also a few cases where bland messages such
2207 as "unrouteable address" or "local delivery error" are given.
2209 PH/43 $value is now also set for the "else" part of a ${run expansion.
2211 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2212 being worked on, but at least Exim now implements the latest version to
2215 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2216 process to complete was treated as a delivery failure, a timeout while
2217 writing the message to the pipe was logged, but erroneously treated as a
2218 successful delivery. Such timeouts include transport filter timeouts. For
2219 consistency with the overall process timeout, these timeouts are now
2220 treated as errors, giving rise to delivery failures by default. However,
2221 there is now a new Boolean option for the pipe transport called
2222 timeout_defer, which, if set TRUE, converts the failures into defers for
2223 both kinds of timeout. A transport filter timeout is now identified in
2226 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2227 systems where "make" and "gmake" are different, calling "gmake" at top
2228 level broke things. I've arranged for the value of $(MAKE) to be passed
2229 from the Makefile to this script so that it can call the same version of
2233 A note about Exim versions 4.44 and 4.50
2234 ----------------------------------------
2236 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2237 changes of various kinds. As a consequence, a big documentation update was
2238 needed. This delayed the release for rather longer than seemed good, especially
2239 in the light of a couple of (minor) security issues. Therefore, the changes
2240 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2241 release. So 4.44 and 4.50 are in effect two different branches that both start
2244 I have left the 4.50 change log unchanged; it contains all the changes since
2245 4.43. The change log for 4.44 is below; many of its items are identical to
2246 those for 4.50. This seems to be the most sensible way to preserve the
2247 historical information.
2253 1. Minor wording change to the doc/README.SIEVE file.
2255 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2256 computation of the current number of files was incorrect.
2258 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2259 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2262 4. Give more explanation in the error message when the command for a transport
2263 filter fails to execute.
2265 5. There are several places where Exim runs a non-Exim command in a
2266 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2267 was being done only for the command run by the queryprogram router. It is
2268 now done for all such subprocesses. The other cases are: ${run, transport
2269 filters, and the commands run by the lmtp and pipe transports.
2271 6. Added CONFIGURE_GROUP build-time option.
2273 7. Some older OS have a limit of 256 on the maximum number of file
2274 descriptors. Exim was using setrlimit() to set 1000 as a large value
2275 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2276 systems. I've change it so that if it can't get 1000, it tries for 256.
2278 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2279 was an oversight, and furthermore, ever since the addition of extra
2280 controls (e.g. 4.43/32), the checks on when to allow different forms of
2281 "control" were broken. There should now be diagnostics for all cases when a
2282 control that does not make sense is encountered.
2284 9. Added the /retain_sender option to "control=submission".
2286 10. $recipients is now available in the predata ACL (oversight).
2288 11. Tidy the search cache before the fork to do a delivery from a message
2289 received from the command line. Otherwise the child will trigger a lookup
2290 failure and thereby defer the delivery if it tries to use (for example) a
2291 cached ldap connection that the parent has called unbind on.
2293 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2294 of $address_data from the recipient verification was clobbered by the
2295 sender verification.
2297 13. The value of address_data from a sender verification is now available in
2298 $sender_address_data in subsequent conditions in the ACL statement.
2300 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2302 15. Added a new option "connect=<time>" to callout options, to set a different
2305 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2306 was its contents. (It was OK if the option was not defined at all.)
2308 17. A "Completed" log line is now written for messages that are removed from
2309 the spool by the -Mrm option.
2311 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2312 information about exactly what failed.
2314 19. Added -dd to debug only the daemon process.
2316 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2317 handles timeouts, both on the server side and network timeouts. Renamed the
2318 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2320 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2321 It is now set to "smtps".
2323 22. $host_address is now set to the target address during the checking of
2324 ignore_target_hosts.
2326 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2327 being passed; this would have caused $sender_host_name to have been used if
2328 matching the list had actually called for a host name (not very likely,
2329 since this list is usually IP addresses). A host name is now passed as
2332 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2333 code that specifies a non-restarting handler (typically sigaction() in
2334 modern systems) in an attempt to fix a rare and obscure crash bug.
2336 25. Narrowed the window for a race in the daemon that could cause it to ignore
2337 SIGCHLD signals. This is not a major problem, because they are used only to
2338 wake it up if nothing else does.
2340 26. A malformed maildirsize file could cause Exim to calculate negative values
2341 for the mailbox size or file count. Odd effects could occur as a result.
2342 The maildirsize information is now recalculated if the size or filecount
2345 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2346 support for a long time. Removed HAVE_SYS_VFS_H.
2348 28. Installed the latest version of exipick from John Jetmore.
2350 29. In an address list, if the pattern was not a regular expression, an empty
2351 subject address (from a bounce message) matched only if the pattern was an
2352 empty string. Non-empty patterns were not even tested. This was the wrong
2353 because it is perfectly reasonable to use an empty address as part of a
2354 database query. An empty address is now tested by patterns that are
2355 lookups. However, all the other forms of pattern expect the subject to
2356 contain a local part and a domain, and therefore, for them, an empty
2357 address still always fails if the pattern is not itself empty.
2359 30. Exim went into a mad DNS loop when attempting to do a callout where the
2360 host was specified on an smtp transport, and looking it up yielded more
2361 than one IP address.
2363 31. Re-factored the code for checking spool and log partition space into a
2364 function that finds that data and another that does the check. The former
2365 is then used to implement four new variables: $spool_space, $log_space,
2366 $spool_inodes, and $log_inodes.
2368 32. The RFC2047 encoding function was originally intended for short strings
2369 such as real names; it was not keeping to the 75-character limit for
2370 encoded words that the RFC imposes. It now respects the limit, and
2371 generates multiple encoded words if necessary. To be on the safe side, I
2372 have increased the buffer size for the ${rfc2047: expansion operator from
2375 33. It is now permitted to omit both strings after an "if" condition; if the
2376 condition is true, the result is "true". As before, when the second string
2377 is omitted, a false condition yields an empty string. This makes it less
2378 cumbersome to write custom ACL and router conditions.
2380 34. Failure to deliver a bounce message always caused it to be frozen, even if
2381 there was an errors_to setting on the router. The errors_to setting is now
2384 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2385 canonical form (fully expanded) before being placed in
2386 $sender_host_address.
2388 36. The table in the code that translates DNS record types into text (T_A to
2389 "A" for instance) was missing entries for NS and CNAME. It is just possible
2390 that this could have caused confusion if both these types were looked up
2391 for the same domain, because the text type is used as part of Exim's
2392 per-process caching. But the chance of anyone hitting this buglet seems
2395 37. The dnsdb lookup has been extended in a number of ways.
2397 (1) There is a new type, "zns", which walks up the domain tree until it
2398 finds some nameserver records. It should be used with care.
2400 (2) There is a new type, "mxh", which is like "mx" except that it returns
2401 just the host names, not the priorities.
2403 (3) It is now possible to give a list of domains (or IP addresses) to be
2404 looked up. The behaviour when one of the lookups defers can be
2405 controlled by a keyword.
2407 (4) It is now possible to specify the separator character for use when
2408 multiple records are returned.
2410 38. The dnslists ACL condition has been extended: it is now possible to supply
2411 a list of IP addresses and/or domains to be looked up in a particular DNS
2414 39. Added log_selector=+queue_time_overall.
2416 40. When running the queue in the test harness, wait just a tad after forking a
2417 delivery process, to get repeatability of debugging output.
2419 41. Include certificate and key file names in error message when GnuTLS fails
2420 to set them up, because the GnuTLS error message doesn't include the name
2421 of the failing file when there is a problem reading it.
2423 42. Allow both -bf and -bF in the same test run.
2425 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2427 44. The "Exiscan patch" is now merged into the mainline Exim source.
2429 45. Sometimes the final signoff response after QUIT could fail to get
2430 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2431 < 0 before doing a fflush(). This bug looks as though it goes back to the
2432 introduction of TLS in release 3.20, but "sometimes" must have been rare
2433 because the tests only now provoked it.
2435 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2436 (this can affect the format of dates).
2438 47. exim_tidydb, when checking for the continued existence of a message for
2439 which it has found a message-specific retry record, was not finding
2440 messages that were in split spool directories. Consequently, it was
2441 deleting retry records that should have stayed in existence.
2443 48. Steve fixed some bugs in eximstats.
2445 49. The SPA authentication driver was not abandoning authentication and moving
2446 on to the next authenticator when an expansion was forced to fail,
2447 contradicting the general specification for all authenticators. Instead it
2448 was generating a temporary error. It now behaves as specified.
2450 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2451 (the order specifies the preference for clients). The order is now AES256,
2452 AES128, 3DES, ARCFOUR128.
2454 51. Small patch to Sieve code - explicitly set From: when generating an
2457 52. Exim crashed if a remote delivery caused a very long error message to be
2458 recorded - for instance if somebody sent an entire SpamAssassin report back
2459 as a large number of 550 error lines. This bug was coincidentally fixed by
2460 increasing the size of one of Exim's internal buffers (big_buffer) that
2461 happened as part of the Exiscan merge. However, to be on the safe side, I
2462 have made the code more robust (and fixed the comments that describe what
2465 53. Now that there can be additional text after "Completed" in log lines (if
2466 the queue_time_overall log selector is set), a one-byte patch to exigrep
2467 was needed to allow it to recognize "Completed" as not the last thing in
2470 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2471 patch that reportedly fixes this has been added. I am not expert enough to
2472 create a test for it. This is what the patch creator wrote:
2474 "I found a little strange behaviour of ldap code when working with
2475 Windows 2003 AD Domain, where users was placed in more than one
2476 Organization Units. When I tried to give exim partial DN, the exit code
2477 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2478 But simultaneously result of request was absolutely normal ldap result,
2479 so I produce this patch..."
2481 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2482 REFERENCE, so I have modified the code to exclude the patch when that macro
2485 55. Some experimental protocols are using DNS PTR records for new purposes. The
2486 keys for these records are domain names, not reversed IP addresses. The
2487 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2488 leaves it alone. Component reversal etc. now happens only for IP addresses.
2491 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2493 57. Double the size of the debug message buffer (to 2048) so that more of very
2494 long debug lines gets shown.
2496 58. The exicyclog utility now does better if the number of log files to keep
2497 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2499 59. Two changes related to the smtp_active_hostname option:
2501 (1) $smtp_active_hostname is now available as a variable.
2502 (2) The default for smtp_banner uses $smtp_active_hostname instead
2503 of $primary_hostname.
2505 60. The host_aton() function is supposed to be passed a string that is known
2506 to be a valid IP address. However, in the case of IPv6 addresses, it was
2507 not checking this. This is a hostage to fortune. Exim now panics and dies
2508 if the condition is not met. A case was found where this could be provoked
2509 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2510 components; fortuitously, this particular loophole had already been fixed
2511 by change 4.50/55 above.
2513 If there are any other similar loopholes, the new check in host_aton()
2514 itself should stop them being exploited. The report I received stated that
2515 data on the command line could provoke the exploit when Exim was running as
2516 exim, but did not say which command line option was involved. All I could
2517 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2518 running as the user.
2521 61. There was a buffer overflow vulnerability in the SPA authentication code
2522 (which came originally from the Samba project). I have added a test to the
2523 spa_base64_to_bits() function which I hope fixes it.
2526 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2527 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2529 63. The daemon start-up calls getloadavg() while still root for those OS that
2530 need the first call to be done as root, but it missed one case: when
2531 deliver_queue_load_max is set with deliver_drop_privilege. This is
2532 necessary for the benefit of the queue runner, because there is no re-exec
2533 when deliver_drop_privilege is set.
2535 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2536 This has been fixed.
2538 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2539 was in use, was not putting the data itself into the right store pool;
2540 consequently, it could be overwritten for a subsequent message in the same
2541 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2544 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2546 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2547 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2548 the function were treating the return as a boolean value, which happened to
2549 work because 0=false and not-0=true, but is not correct code.
2551 68. The host_aton() function was not handling scoped IPv6 addresses (those
2552 with, for example, "%eth0" on the end) correctly.
2554 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2555 negated items (that is, ~something) in unsigned ints. Some compilers
2556 apparently mutter when there is no cast.
2558 70. If an address verification called from an ACL failed, and did not produce a
2559 user-specific message (i.e. there was only a "system" message), nothing was
2560 put in $acl_verify_message. In this situation, it now puts the system
2563 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2564 session; change 4.31/43 added the unwanted input to the log line - except
2565 that it did not do this in the start of session case. It now does.
2567 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2568 This is wrong and can cause the other end to generate a synchronization
2569 error if it is another Exim or anything else that does the synchronization
2570 check. A QUIT command is no longer sent after a timeout.
2572 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2573 during host lookups.
2575 74. The defer_ok option of callout verification was not working if it was used
2576 when verifying addresses in header lines, that is, for this case:
2578 verify = header_sender/callout=defer_ok
2580 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2581 those file descriptors could be used for SMTP connections. If anything
2582 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2583 could be sent to the SMTP client, causing chaos. The daemon now opens
2584 stdin, stdout, and stderr to /dev/null when it puts itself into the
2587 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2588 log by default. The user can override this with suitable Perl magic.
2590 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2591 the log message when discard triggers, was not working for the DATA ACL or
2592 for the non-SMTP ACL.
2594 78. Error message wording change in sieve.c.
2596 79. If smtp_accept_max_per_host was set, the number of connections could be
2597 restricted to fewer than expected, because the daemon was trying to set up
2598 a new connection before checking whether the processes handling previous
2599 connections had finished. The check for completed processes is now done
2600 earlier. On busy systems, this bug wouldn't be noticed because something
2601 else would have woken the daemon, and it would have reaped the completed
2604 80. If a message was submitted locally by a user whose login name contained one
2605 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2606 It caused a spool format error. I have fixed the spool reading code. A
2607 related problem was that the "from" clause in the Received: line became
2608 illegal because of the space(s). It is now covered by ${quote_local_part.
2610 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2613 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2615 83. Patch from Sieve maintainer to fix unterminated string problem in
2616 "vacation" handling.
2618 84. Some minor changes to the Linux configuration files to help with other
2619 OS variants using glibc.
2621 85. One more patch for Sieve to update vacation handling to latest spec.
2624 ----------------------------------------------------
2625 See the note above about the 4.44 and 4.50 releases.
2626 ----------------------------------------------------
2632 1. Change 4.43/35 introduced a bug that caused file counts to be
2633 incorrectly computed when quota_filecount was set in an appendfile
2636 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2637 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2640 3. Give more explanation in the error message when the command for a transport
2641 filter fails to execute.
2643 4. There are several places where Exim runs a non-Exim command in a
2644 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2645 was being done only for the command run by the queryprogram router. It is
2646 now done for all such subprocesses. The other cases are: ${run, transport
2647 filters, and the commands run by the lmtp and pipe transports.
2649 5. Some older OS have a limit of 256 on the maximum number of file
2650 descriptors. Exim was using setrlimit() to set 1000 as a large value
2651 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2652 systems. I've change it so that if it can't get 1000, it tries for 256.
2654 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2655 was an oversight, and furthermore, ever since the addition of extra
2656 controls (e.g. 4.43/32), the checks on when to allow different forms of
2657 "control" were broken. There should now be diagnostics for all cases when a
2658 control that does not make sense is encountered.
2660 7. $recipients is now available in the predata ACL (oversight).
2662 8. Tidy the search cache before the fork to do a delivery from a message
2663 received from the command line. Otherwise the child will trigger a lookup
2664 failure and thereby defer the delivery if it tries to use (for example) a
2665 cached ldap connection that the parent has called unbind on.
2667 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2668 of $address_data from the recipient verification was clobbered by the
2669 sender verification.
2671 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2672 was its contents. (It was OK if the option was not defined at all.)
2674 11. A "Completed" log line is now written for messages that are removed from
2675 the spool by the -Mrm option.
2677 12. $host_address is now set to the target address during the checking of
2678 ignore_target_hosts.
2680 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2681 being passed; this would have caused $sender_host_name to have been used if
2682 matching the list had actually called for a host name (not very likely,
2683 since this list is usually IP addresses). A host name is now passed as
2686 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2687 code that specifies a non-restarting handler (typically sigaction() in
2688 modern systems) in an attempt to fix a rare and obscure crash bug.
2690 15. Narrowed the window for a race in the daemon that could cause it to ignore
2691 SIGCHLD signals. This is not a major problem, because they are used only to
2692 wake it up if nothing else does.
2694 16. A malformed maildirsize file could cause Exim to calculate negative values
2695 for the mailbox size or file count. Odd effects could occur as a result.
2696 The maildirsize information is now recalculated if the size or filecount
2699 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2700 support for a long time. Removed HAVE_SYS_VFS_H.
2702 18. Updated exipick to current release from John Jetmore.
2704 19. Allow an empty sender to be matched against a lookup in an address list.
2705 Previously the only cases considered were a regular expression, or an
2708 20. Exim went into a mad DNS lookup loop when doing a callout where the
2709 host was specified on the transport, if the DNS lookup yielded more than
2712 21. The RFC2047 encoding function was originally intended for short strings
2713 such as real names; it was not keeping to the 75-character limit for
2714 encoded words that the RFC imposes. It now respects the limit, and
2715 generates multiple encoded words if necessary. To be on the safe side, I
2716 have increased the buffer size for the ${rfc2047: expansion operator from
2719 22. Failure to deliver a bounce message always caused it to be frozen, even if
2720 there was an errors_to setting on the router. The errors_to setting is now
2723 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2724 canonical form (fully expanded) before being placed in
2725 $sender_host_address.
2727 24. Updated eximstats to version 1.33
2729 25. Include certificate and key file names in error message when GnuTLS fails
2730 to set them up, because the GnuTLS error message doesn't include the name
2731 of the failing file when there is a problem reading it.
2733 26. Expand error message when OpenSSL has problems setting up cert/key files.
2736 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2737 (this can affect the format of dates).
2739 28. exim_tidydb, when checking for the continued existence of a message for
2740 which it has found a message-specific retry record, was not finding
2741 messages that were in split spool directories. Consequently, it was
2742 deleting retry records that should have stayed in existence.
2744 29. eximstats updated to version 1.35
2745 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2746 1.35 - bugfix such that pie charts by volume are generated correctly
2748 30. The SPA authentication driver was not abandoning authentication and moving
2749 on to the next authenticator when an expansion was forced to fail,
2750 contradicting the general specification for all authenticators. Instead it
2751 was generating a temporary error. It now behaves as specified.
2753 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2754 (the order specifies the preference for clients). The order is now AES256,
2755 AES128, 3DES, ARCFOUR128.
2757 31. Small patch to Sieve code - explicitly set From: when generating an
2760 32. Exim crashed if a remote delivery caused a very long error message to be
2761 recorded - for instance if somebody sent an entire SpamAssassin report back
2762 as a large number of 550 error lines. This bug was coincidentally fixed by
2763 increasing the size of one of Exim's internal buffers (big_buffer) that
2764 happened as part of the Exiscan merge. However, to be on the safe side, I
2765 have made the code more robust (and fixed the comments that describe what
2768 33. Some experimental protocols are using DNS PTR records for new purposes. The
2769 keys for these records are domain names, not reversed IP addresses. The
2770 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2771 leaves it alone. Component reversal etc. now happens only for IP addresses.
2774 34. The host_aton() function is supposed to be passed a string that is known
2775 to be a valid IP address. However, in the case of IPv6 addresses, it was
2776 not checking this. This is a hostage to fortune. Exim now panics and dies
2777 if the condition is not met. A case was found where this could be provoked
2778 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2779 components; fortuitously, this particular loophole had already been fixed
2780 by change 4.50/55 or 4.44/33 above.
2782 If there are any other similar loopholes, the new check in host_aton()
2783 itself should stop them being exploited. The report I received stated that
2784 data on the command line could provoke the exploit when Exim was running as
2785 exim, but did not say which command line option was involved. All I could
2786 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2787 running as the user.
2790 35. There was a buffer overflow vulnerability in the SPA authentication code
2791 (which came originally from the Samba project). I have added a test to the
2792 spa_base64_to_bits() function which I hope fixes it.
2795 36. The daemon start-up calls getloadavg() while still root for those OS that
2796 need the first call to be done as root, but it missed one case: when
2797 deliver_queue_load_max is set with deliver_drop_privilege. This is
2798 necessary for the benefit of the queue runner, because there is no re-exec
2799 when deliver_drop_privilege is set.
2801 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2802 was in use, was not putting the data itself into the right store pool;
2803 consequently, it could be overwritten for a subsequent message in the same
2804 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2807 38. Sometimes the final signoff response after QUIT could fail to get
2808 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2809 < 0 before doing a fflush(). This bug looks as though it goes back to the
2810 introduction of TLS in release 3.20, but "sometimes" must have been rare
2811 because the tests only now provoked it.
2817 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2818 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2819 Now it returns an integer. A number of calls were still expecting a T/F
2820 return. Fortuitously, in all cases, the tests worked in OK situations,
2821 which is the norm. However, things would have gone wrong on any write
2822 failures on the smtp file descriptor. This function is used when sending
2823 messages over SMTP and also when doing verify callouts.
2825 2. When Exim is called to do synchronous delivery of a locally submitted
2826 message (the -odf or -odi options), it no longer closes stderr before doing
2829 3. Implemented the mua_wrapper option.
2831 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2833 5. Implemented the functions header_remove(), header_testname(),
2834 header_add_at_position(), and receive_remove_recipient(), and exported them
2837 6. If an ACL "warn" statement specified the addition of headers, Exim already
2838 inserted X-ACL-Warn: at the start if there was no header name. However, it
2839 was not making this test for the second and subsequent header lines if
2840 there were newlines in the string. This meant that an invalid header could
2841 be inserted if Exim was badly configured.
2843 7. Allow an ACL "warn" statement to add header lines at the start or after all
2844 the Received: headers, as well as at the end.
2846 8. Added the rcpt_4xx retry error code.
2848 9. Added postmaster_mailfrom=xxx to callout verification option.
2850 10. Added mailfrom=xxxx to the callout verification option, for verify=
2853 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2854 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2856 12. Inserted some casts to stop certain compilers warning when using pointer
2857 differences as field lengths or precisions in printf-type calls (mostly
2858 affecting debugging statements).
2860 13. Added optional readline() support for -be (dynamically loaded).
2862 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2863 same clock tick as a message's arrival, so that its received time was the
2864 same as the "first fail" time on the retry record, and that message
2865 remained on the queue past the ultimate address timeout, every queue runner
2866 would try a delivery (because it was past the ultimate address timeout) but
2867 after another failure, the ultimate address timeout, which should have then
2868 bounced the address, did not kick in. This was a "< instead of <=" error;
2869 in most cases the first failure would have been in the next clock tick
2870 after the received time, and all would be well.
2872 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2873 being recognized when the domain list was tested by the match_domain
2874 condition in an expansion string.
2876 16. Added the ${str2b64: operator.
2878 17. Exim was always calling setrlimit() to set a large limit for the number of
2879 processes, without checking whether the existing limit was already
2880 adequate. (It did check for the limit on file descriptors.) Furthermore,
2881 errors from getrlimit() and setrlimit() were being ignored. Now they are
2882 logged to the main and panic logs, but Exim does carry on, to try to do its
2883 job under whatever limits there are.
2885 18. Imported PCRE 5.0.
2887 19. Trivial typo in log message " temporarily refused connection" (the leading
2890 20. If the log selector return_path_on_delivery was set and an address was
2891 redirected to /dev/null, the delivery process crashed because it assumed
2892 that a return path would always be set for a "successful" delivery. In this
2893 case, the whole delivery is bypassed as an optimization, and therefore no
2896 21. Internal re-arrangement: the function for sending a challenge and reading
2897 a response while authentication was assuming a zero-terminated challenge
2898 string. It's now changed to take a pointer and a length, to allow for
2899 binary data in such strings.
2901 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2903 23. Exim was not respecting finduser_retries when seeking the login of the
2904 uid under which it was called; it was always trying 10 times. (The default
2905 setting of finduser_retries is zero.) Also, it was sleeping after the final
2906 failure, which is pointless.
2908 24. Implemented tls_on_connect_ports.
2910 25. Implemented acl_smtp_predata.
2912 26. If the domain in control=submission is set empty, Exim assumes that the
2913 authenticated id is a complete email address when it generates From: or
2914 Sender: header lines.
2916 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2917 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2918 chown and chgrp in /bin and hostname in /usr/bin.
2920 28. Exim was keeping the "process log" file open after each use, just as it
2921 does for the main log. This opens the possibility of it remaining open for
2922 long periods when the USR1 signal hits a daemon. Occasional processlog
2923 errors were reported, that could have been caused by this. Anyway, it seems
2924 much more sensible not to leave this file open at all, so that is what now
2927 29. The long-running daemon process does not normally write to the log once it
2928 has entered its main loop, and it closes the log before doing so. This is
2929 so that log files can straightforwardly be renamed and moved. However,
2930 there are a couple of unusual error situations where the daemon does write
2931 log entries, and I had neglected to close the log afterwards.
2933 30. The text of an SMTP error response that was received during a remote
2934 delivery was being truncated at 512 bytes. This is too short for some of
2935 the long messages that one sometimes sees. I've increased the limit to
2938 31. It is now possible to make retry rules that apply only when a message has a
2939 specific sender, in particular, an empty sender.
2941 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2942 it possible to be selective about when SMTP synchronization is enforced.
2944 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2946 32. Implemented hosts_connection_nolog.
2948 33. Added an ACL for QUIT.
2950 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2953 35. Added mailbox_size and mailbox_filecount to appendfile.
2955 36. Added control = no_multiline_responses to ACLs.
2957 37. There was a bug in the logic of the code that waits for the clock to tick
2958 in the case where the clock went backwards by a substantial amount such
2959 that the microsecond fraction of "now" was more than the microsecond
2960 fraction of "then" (but the whole seconds number was less).
2962 38. Added support for the libradius Radius client library this is found on
2963 FreeBSD (previously only the radiusclient library was supported).
2969 1. When certain lookups returned multiple values in the form name=value, the
2970 quoting of the values was not always being done properly. Specifically:
2971 (a) If the value started with a double quote, but contained no whitespace,
2973 (b) If the value contained whitespace other than a space character (i.e.
2974 tabs or newlines or carriage returns) it was not quoted.
2975 This fix has been applied to the mysql and pgsql lookups by writing a
2976 separate quoting function and calling it from the lookup code. The fix
2977 should probably also be applied to nisplus, ibase and oracle lookups, but
2978 since I cannot test any of those, I have not disturbed their existing code.
2980 2. A hit in the callout cache for a specific address caused a log line with no
2981 reason for rejecting RCPT. Now it says "Previous (cached) callout
2982 verification failure".
2984 3. There was an off-by-one bug in the queryprogram router. An over-long
2985 return line was truncated at 256 instead of 255 characters, thereby
2986 overflowing its buffer with the terminating zero. As well as fixing this, I
2987 have increased the buffer size to 1024 (and made a note to document this).
2989 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2990 when Exim is waiting for an SMTP response from a remote server, Exim
2991 restarts its select() call on the socket, thereby resetting its timeout.
2992 This is not a problem when such interrupts are rare. Somebody set up a cron
2993 job to run exiwhat every 2 minutes, which is less than the normal select()
2994 timeout (5 or 10 minutes). This meant that the select() timeout never
2995 kicked in because it was always reset. I have fixed this by comparing the
2996 time when an interrupt arrives with the time at the start of the first call
2997 to select(). If more time than the timeout has elapsed, the interrupt is
2998 treated as a timeout.
3000 5. Some internal re-factoring in preparation for the addition of Sieve
3001 extensions (by MH). In particular, the "personal" test is moved to a
3002 separate function, and given an option for scanning Cc: and Bcc: (which is
3003 not set for Exim filters).
3005 6. When Exim created an email address using the login of the caller as the
3006 local part (e.g. when creating a From: or Sender: header line), it was not
3007 quoting the local part when it contained special characters such as @.
3009 7. Installed new OpenBSD configuration files.
3011 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3012 try to make them clearer.
3014 9. Callout options, other than the timeout value, were being ignored when
3015 verifying sender addresses in header lines. For example, when using
3017 verify = header_sender/callout=no_cache
3019 the cache was (incorrectly) being used.
3021 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3022 only the Cygwin environment.
3024 11. When return_path_on_delivery was set as a log selector, if different remote
3025 addresses in the same message used different return paths and parallel
3026 remote delivery occurred, the wrong values would sometimes be logged.
3027 (Whenever a remote delivery process finished, the return path value from
3028 the most recently started remote delivery process was logged.)
3030 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3031 lines when AUTH and/or TLS are in use. This is the "received protocol"
3032 field. Exim used to use "asmtp" for authenticated SMTP, without any
3033 indication (in the protocol name) for TLS use. Now it follows the RFC and
3034 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3035 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3036 names appear in log lines as well as in Received: header lines.
3038 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3039 extensions, and comparison tests, and to fix some bugs.
3041 14. Changes to the "personal" filter test:
3043 (1) The test was buggy in that it was just doing the equivalent of
3044 "contains" tests on header lines. For example, if a user's address was
3045 anne@some.where, the "personal" test would incorrectly be true for
3047 To: susanne@some.where
3049 This test is now done by extracting each address from the header in turn,
3050 and checking the entire address. Other tests that are part of "personal"
3051 are now done using regular expressions (for example, to check local parts
3052 of addresses in From: header lines).
3054 (2) The list of non-personal local parts in From: addresses has been
3055 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3056 taken from the Sieve specification recommendations.
3058 (3) If the message contains any header line starting with "List-" it is
3059 treated as non-personal.
3061 (4) The test for "circular" in the Subject: header line has been removed
3062 because it now seems ill-conceived.
3064 15. Minor typos in src/EDITME comments corrected.
3066 16. Installed latest exipick from John Jetmore.
3068 17. If headers_add on a router specified a text string that was too long for
3069 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3070 of string_sprintf() is now avoided.
3072 18. $message_body_size was not set (it was always zero) when running the DATA
3073 ACL and the local_scan() function.
3075 19. For the "mail" command in an Exim filter, no default was being set for
3076 the once_repeat time, causing a random time value to be used if "once" was
3077 specified. (If the value happened to be <= 0, no repeat happened.) The
3078 default is now 0s, meaning "never repeat". The "vacation" command was OK
3079 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3080 (I found it when inspecting the code).
3082 20. There is now an overall timeout for performing a callout verification. It
3083 defaults to 4 times the callout timeout, which applies to individual SMTP
3084 commands during the callout. The overall timeout applies when there is more
3085 than one host that can be tried. The timeout is checked before trying the
3086 next host. This prevents very long delays if there are a large number of
3087 hosts and all are timing out (e.g. when the network connections are timing
3088 out). The value of the overall timeout can be changed by specifying an
3089 additional sub-option for "callout", called "maxwait". For example:
3091 verify = sender/callout=5s,maxwait=20s
3093 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3094 to the end before writing, but this should make it even safer).
3096 22. Exim was forgetting that it had advertised PIPELINING for the second and
3097 subsequent messages on an SMTP connection. It was also not resetting its
3098 memory on STARTTLS and an internal HELO.
3100 23. When Exim logs an SMTP synchronization error within a session, it now
3101 records whether PIPELINING has been advertised or not.
3103 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3104 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3105 rather than long int.
3107 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3109 26. Added the never_mail option to autoreply.
3115 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3116 crash if the getsockname() call failed; this can happen if a connection is
3117 closed very soon after it is established. The problem was simply in the
3118 order in which certain operations were done, causing Exim to try to write
3119 to the SMTP stream before it had set up the file descriptor. The bug has
3120 been fixed by making things happen in the correct order.
3126 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3127 before the connection was closed, thus losing the rejection response.
3129 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3130 some early Solaris releases, but causes trouble in current releases where
3131 socklen_t is defined.
3133 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3136 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3139 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3140 that releasing the top of it at the end releases what was used for sub-
3141 expansions (unless the block got too big). However, discard this block if
3142 the first thing is a variable or header, so that we can use its block when
3143 it is dynamic (useful for very large $message_headers, for example).
3145 6. Lookups now cache *every* query, not just the most recent. A new, separate
3146 store pool is used for this. It can be recovered when all lookup caches are
3147 flushed. Lookups now release memory at the end of their result strings.
3148 This has involved some general refactoring of the lookup sources.
3150 7. Some code has been added to the store_xxx() functions to reduce the amount
3151 of flapping under certain conditions.
3153 8. log_incoming_interface used to affect only the <= reception log lines. Now
3154 it causes the local interface and port to be added to several more SMTP log
3155 lines, for example "SMTP connection from", and rejection lines.
3157 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3159 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3161 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3162 could be overwritten at the end of the current message (or the start of a
3163 new message if it was set in a HELO ACL). The value is now preserved for
3164 the duration of the SMTP connection.
3166 12. If a transport had a headers_rewrite setting, and a matching header line
3167 contained an unqualified address, that address was qualified, even if it
3168 did not match any rewriting rules. The underlying bug was that the values
3169 of the flags that permit the existence of unqualified sender and recipient
3170 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3171 non-local messages, and by -bnq for local messages) were not being
3172 preserved with the message after it was received.
3174 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3175 "next input=" as part of the text comprising the host identity instead of
3176 the correct text. The code was using the same buffer for two different
3177 strings. However, depending on which order the printing function evaluated
3178 its arguments, the bug did not always show up. Under Linux, for example, my
3179 test suite worked just fine.
3181 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3182 doesn't work with some older versions of Perl. It has been changed to "my",
3183 which in any case is probably the better facility to use.
3185 15. A really picky compiler found some instances of statements for creating
3186 error messages that either had too many or two few arguments for the format
3189 16. The size of the buffer for calls to the DNS resolver has been increased
3190 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3191 for addresses that have a lot of PTR records. This alleviates a problem; it
3192 does not fully solve it.
3194 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3195 buffer now truncates the list and logs the incident, which is the same
3196 action as happens when Exim is looking up a host name and its aliases.
3197 Previously in this situation something unpredictable would happen;
3198 sometimes it was "internal error: store_reset failed".
3200 18. If a server dropped the connection unexpectedly when an Exim client was
3201 using GnuTLS and trying to read a response, the client delivery process
3202 crashed while trying to generate an error log message.
3204 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3205 string, for example:
3207 warn message = H1: something\nH2: something
3209 the text was added as a single header line from Exim's point of view
3210 though it ended up OK in the delivered message. However, searching for the
3211 second and subsequent header lines using $h_h2: did not work. This has been
3212 fixed. Similarly, if a system filter added multiple headers in this way,
3213 the routers could not see them.
3215 20. Expanded the error message when iplsearch is called with an invalid key to
3216 suggest using net-iplsearch in a host list.
3218 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3219 ACLs are no longer actually imposed (and a message to that effect is
3222 22. If a "gecos" field in a passwd entry contained escaped characters, in
3223 particular, if it contained a \" sequence, Exim got it wrong when building
3224 a From: or a Sender: header from that name. A second bug also caused
3225 incorrect handling when an unquoted " was present following a character
3226 that needed quoting.
3228 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3229 was not being matched caselessly.
3231 24. Arranged for all hyphens in the exim.8 source to be escaped with
3234 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3235 properly. Recipient callout cache records were still being keyed to include
3236 the sender, even when use_sender was set false. This led to far more
3237 callouts that were necessary. The sender is no longer included in the key
3238 when use_sender is false.
3240 26. Added "control = submission" modifier to ACLs.
3242 27. Added the ${base62d: operator to decode base 62 numbers.
3244 28. dnsdb lookups can now access SRV records.
3246 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3247 the configuration file.
3249 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3250 (-v) mode. This makes the output for a verbose queue run more intelligible.
3252 31. Added a use_postmaster feature to recipient callouts.
3254 32. Added the $body_zerocount variable, containing the number of binary zero
3255 bytes in the message body.
3257 33. The time of last modification of the "new" subdirectory is now used as the
3258 "mailbox time last read" when there is a quota error for a maildir
3261 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3263 35. Added +ignore_unknown as a special item in host lists.
3265 36. Code for decoding IPv6 addresses in host lists is now included, even if
3266 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3267 address was recognized as an IP address, but was then not correctly decoded
3268 into binary, causing unexpected and incorrect effects when compared with
3275 1. Very minor rewording of debugging text in manualroute to say "list of
3276 hosts" instead of "hostlist".
3278 2. If verify=header_syntax was set, and a header line with an unqualified
3279 address (no domain) and a large number of spaces between the end of the
3280 name and the colon was received, the reception process suffered a buffer
3281 overflow, and (when I tested it) crashed. This was caused by some obsolete
3282 code that should have been removed. The fix is to remove it!
3284 3. When running in the test harness, delay a bit after writing a bounce
3285 message to get a bit more predictability in the log output.
3287 4. Added a call to search_tidyup() just before forking a reception process. In
3288 theory, someone could use a lookup in the expansion of smtp_accept_max_
3289 per_host which, without the tidyup, could leave open a database connection.
3291 5. Added the variables $recipient_data and $sender_data which get set from a
3292 lookup success in an ACL "recipients" or "senders" condition, or a router
3293 "senders" option, similar to $domain_data and $local_part_data.
3295 6. Moved the writing of debug_print from before to after the "senders" test
3298 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3299 problems for message scanning, either using a data ACL, or using
3300 local_scan() because the Received: header was not generated till after they
3301 were called (in order to set the time as the time of reception completion).
3302 I have revised the way this works. The header is now generated after the
3303 body is received, but before the ACL or local_scan() are called. After they
3304 are run, the timestamp in the header is updated.
3310 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3311 before starting a queue runner without re-exec. This happened only when
3312 deliver_drop_privilege was set or when the Exim user was set to root. The
3313 effect of the bug was that timeouts during subsequent deliveries caused
3314 crashes instead of being properly handled. The handler is now left at its
3315 default (and expected) setting.
3317 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3318 message, again when deliver_drop_privilege is set or Exim is run as root.
3319 The bug described in (1) was not present in this case, but the tidying up
3320 of the other signals was missing. I have made the two cases consistent.
3322 3. The ignore_target_hosts setting on a manualroute router was being ignored
3323 for hosts that were looked up using the /MX notation.
3325 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3328 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3329 operated on the sender address. After changing the $sender_address to <>
3330 for the sender address verify, Exim was re-instated it as the original
3331 (before rewriting) address, but remembering that it had rewritten it, so it
3332 wasn't rewriting it again. This bug also had the effect of breaking the
3333 sender address verification caching when the sender address was rewritten.
3335 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3336 This has been changed so that if the ip literal address matches
3337 ignore_target_hosts, the router declines.
3339 7. Added expansion conditions match_domain, match_address, and match_local_
3340 part (NOT match_host).
3342 8. The placeholder for the Received: header didn't have a length field set.
3344 9. Added code to Exim itself and to exim_lock to test for a specific race
3345 condition that could lead to file corruption when using MBX delivery. The
3346 issue is with the lockfile that is created in /tmp. If this file is removed
3347 after a process has opened it but before that process has acquired a lock,
3348 there is the potential for a second process to recreate the file and also
3349 acquire a lock. This could lead to two Exim processes writing to the file
3350 at the same time. The added code performs the same test as UW imapd; it
3351 checks after acquiring the lock that its file descriptor still refers to
3352 the same named file.
3354 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3355 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3358 11. Added the smtp_active_hostname option. If used, this will typically be made
3359 to depend on the incoming interface address. Because $interface_address is
3360 not set up until the daemon has forked a reception process, error responses
3361 that can happen earlier (such as "too many connections") no longer contain
3364 12. If an expansion in a condition on a "warn" statement fails because a lookup
3365 defers, the "warn" statement is abandoned, and the next ACL statement is
3366 processed. Previously this caused the whole ACL to be aborted.
3368 13. Added the iplsearch lookup type.
3370 14. Added ident_timeout as a log selector.
3372 15. Added tls_certificate_verified as a log selector.
3374 16. Added a global option tls_require_ciphers (compare the smtp transport
3375 option of the same name). This controls incoming TLS connections.
3377 17. I finally figured out how to make tls_require_ciphers do a similar thing
3378 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3379 before starting the TLS session.
3381 18. Tabs are now shown as \t in -bP output.
3383 19. If the log selector return_path_on_delivery was set, Exim crashed when
3384 bouncing a message because it had too many Received: header lines.
3386 20. If two routers both had headers_remove settings, and the first one included
3387 a superfluous trailing colon, the final name in the first list and the
3388 first name in the second list were incorrectly joined into one item (with a
3389 colon in the middle).
3395 1. Added -C and -D options to the exinext utility, mainly to make it easier
3396 to include in the automated testing, but these could be helpful when
3397 multiple configurations are in use.
3399 2. The exinext utility was not formatting the output nicely when there was
3400 an alternate port involved in the retry record key, nor when there was a
3401 message id as well (for retries that were specific to a specific message
3402 and a specific host). It was also confused by IPv6 addresses, because of
3403 the additional colons they contain. I have fixed the IPv4 problem, and
3404 patched it up to do a reasonable job for IPv6.
3406 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3407 delivery, the log line now contains "pipelined" if PIPELINING was used.
3409 4. An SMTP transport process used to panic and die if the bind() call to set
3410 an explicit outgoing interface failed. This has been changed; it is now
3411 treated in the same way as a connect() failure.
3413 5. A reference to $sender_host_name in the part of a conditional expansion
3414 that was being skipped was still causing a DNS lookup. This no longer
3417 6. The def: expansion condition was not recognizing references to header lines
3418 that used bh_ and bheader_.
3420 7. Added the _cache feature to named lists.
3422 8. The code for checking quota_filecount in the appendfile transport was
3423 allowing one more file than it should have been.
3425 9. For compatibility with Sendmail, the command line option
3433 and sets the incoming protocol and host name (for trusted callers). The
3434 host name and its colon can be omitted when only the protocol is to be set.
3435 Note the Exim already has two private options, -pd and -ps, that refer to
3436 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3437 "s", but I don't think that's a major issue.
3439 10. A number of refactoring changes to the code, none of which should affect
3442 (a) The number of logging options was getting close to filling up the
3443 32-bit word that was used as a bit map. I have split them into two classes:
3444 those that are passed in the argument to log_write(), and those that are
3445 only ever tested independently outside of that function. These are now in
3446 separate 32-bit words, so there is plenty of room for expansion again.
3447 There is no change in the user interface or the logging behaviour.
3449 (b) When building, for example, log lines, the code previously used a
3450 macro that called string_cat() twice, in order to add two strings. This is
3451 not really sufficiently general. Furthermore, there was one instance where
3452 it was actually wrong because one of the argument was used twice, and in
3453 one call a function was used. (As it happened, calling the function twice
3454 did not affect the overall behaviour.) The macro has been replaced by a
3455 function that can join an arbitrary number of extra strings onto a growing
3458 (c) The code for expansion conditions now uses a table and a binary chop
3459 instead of a serial search (which was left over from when there were very
3460 few conditions). Also, it now recognizes conditions like "pam" even when
3461 the relevant support is not compiled in: a suitably worded error message is
3462 given if an attempt is made to use such a condition.
3464 11. Added ${time_interval:xxxxx}.
3466 12. A bug was causing one of the ddress fields not to be passed back correctly
3467 from remote delivery subprocesses. The field in question was not being
3468 subsequently used, so this caused to problems in practice.
3470 13. Added new log selectors queue_time and deliver_time.
3472 14. Might have fixed a bug in maildirsizefile handling that threw up
3473 "unexpected character" debug warnings, and recalculated the data
3474 unnecessarily. In any case, I expanded the warning message to give more
3477 15. Added the message "Restricted characters in address" to the statements in
3478 the default ACL that block characters like @ and % in local parts.
3480 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3481 Three changes have been made:
3483 (a) There was a serious bug; a negative response to MAIL caused the whole
3484 recipient domain to be cached as invalid, thereby blocking all messages
3485 to all local parts at the same domain, from all senders. This bug has
3486 been fixed. The domain is no longer cached after a negative response to
3487 MAIL if the sender used is not empty.
3489 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3492 (c) A new callout option, "use_sender" has been added for people who want
3493 the modified behaviour.
3499 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3502 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3503 indeed breaks things for older releases.
3505 3. Added additional logging to the case where there is a problem reading data
3506 from a filter that is running in a subprocess using a pipe, in order to
3507 try to track down a specific problem.
3509 4. Testing facility fudge: when running in the test harness and attempting
3510 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3511 getting "No route to host". Convert this to a timeout.
3513 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3516 6. Some OS don't have socklen_t but use size_t instead. This affects the
3517 fifth argument of getsockopt() amongst other things. This is now
3518 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3519 can be set for individual OS. I have set it for SunOS5, OSF1, and
3520 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3521 some earlier ones do not.
3523 7. Change 4.30/15 was not doing the test caselessly.
3525 8. The standard form for an IPv6 address literal was being rejected by address
3526 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3527 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3528 this, as well as the form without the "IPv6" on the front (but only when
3529 address literals are enabled, of course).
3531 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3533 10. Exim crashed if a message with an empty sender address specified by -f
3534 encountered a router with an errors_to setting. This could be provoked only
3535 by a command such as
3539 where an empty string was supplied; "<>" did not hit this bug.
3541 11. Installed PCRE release 4.5.
3543 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3544 remained set. It is now erased.
3546 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3547 times from message ids (which are base 36 rather than the normal 62).
3549 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3550 were being counted as actual protocol errors, and logged if the log
3551 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3552 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3553 and DATA following a set of rejected RCPTs do not count as protocol errors.
3554 In other words, Exim assumes they were pipelined, though this may not
3555 actually be the case. Of course, in all cases the client gets an
3556 appropriate error code.
3558 15. If a lookup fails in an ACL condition, a message about the failure may
3559 be available; it is used if testing the ACL cannot continue, because most
3560 such messages specify what the cause of the deferral is. However, some
3561 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3562 that caused an old message to be retained and used if a later statement
3563 caused a defer, replacing the real cause of the deferral.
3565 16. If an IP address had so many PTR records that the DNS lookup buffer
3566 was not large enough to hold them, Exim could crash while trying to process
3567 the truncated data. It now detects and logs this case.
3569 17. Further to 4.21/58, another change has been made: if (and only if) the
3570 first line of a message (the first header line) ends with CRLF, a bare LF
3571 in a subsequent header line has a space inserted after it, so as not to
3572 terminate the header.
3574 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3575 unnecessarily, used atoi() instead of strtol(), and didn't check the
3576 termination when getting file sizes from file names by regex.
3578 19. Completely re-implemented the support for maildirsize files, in the light
3579 of a number of problems with the previous contributed implementation
3580 (4.30/29). In particular:
3582 . If the quota is zero, the maildirsize file is maintained, but no quota is
3585 . If the maildir directory does not exist, it is created before any attempt
3586 to write a maildirsize file.
3588 . The quota value in the file is just a cache; if the quota is changed in
3589 the transport, the new value overrides.
3591 . A regular expression is available for excluding directories from the
3594 20. The autoreply transport checks the characters in options that define the
3595 message's headers; it allows continued headers, but it was checking with
3596 isspace() after an embedded newline instead of explicitly looking for a
3599 21. If all the "regular" hosts to which an address was routed had passed their
3600 expiry times, and had not reached their retry times, the address was
3601 bounced, even if fallback hosts were defined. Now Exim should go on to try
3604 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3605 equivalent code in the SMTP transport. Some hosts send humungous responses
3606 to HELO/EHLO, more than 1024 it seems.
3608 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3609 gives compiler warnings in some environments. I've now done it "properly",
3612 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3613 (because of problems with the built-in one) was declared to return uschar *
3614 instead of char *, causing compiler failure.
3616 25. Fixed a file descriptor leak when processing alias/forward files.
3618 26. Fixed a minor format string issue in dbfn.c.
3620 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3622 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3623 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3626 29. When a sender address is verified, it is cached, to save repeating the test
3627 when there is more than one recipient in a message. However, when the
3628 verification involves a callout, it is possible for different callout
3629 options to be set for different recipients. It is too complicated to keep
3630 track of this in the cache, so now Exim always runs a verification when a
3631 callout is required, relying on the callout cache for the optimization.
3632 The overhead is duplication of the address routing, but this should not be
3635 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3636 to be verified with callout=postmaster, and the main callout worked but the
3637 postmaster check failed, the verification correctly failed. However, if a
3638 subsequent RCPT command asked for sender verification *without* the
3639 postmaster check, incorrect caching caused this verification also to fail,
3642 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3643 it was not caching the DNS options (qualify_single, search_parents) that
3644 were used when the lookup failed. A subsequent lookup with different
3645 options therefore always gave the same answer, though there were cases
3646 where it should not have. (Example: a "domains = !$mx_any" option on a
3647 dnslookup router: the "domains" option is always processed without any
3648 widening, but the router might have qualify_single set.) Now Exim uses the
3649 cached value only when the same options are set.
3651 32. Added John Jetmore's "exipick" utility to the distribution.
3653 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3654 than a timeout (e.g. a certificate is required, and is not provided), an
3655 Exim server now closes the connection immediately. Previously it waited for
3656 the client to close - but if the client is SSL, it seems that they each
3657 wait for each other, leading to a delay before one of them times out.
3659 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3660 maintained 0.8.x compatibility because I don't think many are using it, and
3661 it is clearly obsolete.
3663 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3666 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3667 client certificate was expired. A simple patch fixes this, though I don't
3668 understand the full logic of why the verify callback is called multiple
3671 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3672 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3673 which causes problems with some clients (such as the Certicom SSL Plus
3674 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3675 disables the coutermeasure allowing Eudora to connect."
3677 38. Exim was not checking that a write() to a log file succeeded. This could
3678 lead to Bad Things if a log got too big, in particular if it hit a file
3679 size limit. Exim now panics and dies if it cannot write to a log file, just
3680 as it does if it cannot open a log file.
3682 39. Modified OS/Makefile-Linux so that it now contains
3684 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3686 The two -D definitions ensure that Exim is compiled with large file
3687 support, which makes it possible to handle log files that are bigger than
3690 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3691 instance) a domain was checked against a named list that involved a lookup,
3692 causing $domain_data to be set, then another domain was checked against the
3693 same list, then the first domain was re-checked, the value of $domain_data
3694 after the final check could be wrong. In particular, if the second check
3695 failed, it could be set empty. This bug probably also applied to
3698 41. The strip_trailing_dot option was not being applied to the address given
3699 with the -f command-line option.
3701 42. The code for reading a message's header from the spool was incrementing
3702 $received_count, but never initializing it. This meant that the value was
3703 incorrect (doubled) while delivering a message in the same process in which
3704 it was received. In the most common configuration of Exim, this never
3705 happens - a fresh exec is done - but it can happen when
3706 deliver_drop_privilege is set.
3708 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3709 it now includes up to 150 characters of the unexpected data in the log
3712 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3713 and building data strings. The size of both of these buffers was 10 000
3714 bytes - far larger than anybody would *ever* want, thought I. Needless to
3715 say, somebody hit the limit. I have increased the maximum line length to
3716 20 000 and the maximum data length of concatenated lines to 100 000. I have
3717 also fixed two bugs, because there was no checking on these buffers. Tsk,
3718 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3719 buffer is too small.
3721 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3722 lsearch lookups. Now it does.
3724 46. When parsing a route_list item in a manualroute router, a fixed-length
3725 buffer was used for the list of hosts. I made this 1024 bytes long,
3726 thinking that nobody would ever have a list of hosts that long. Wrong.
3727 Somebody had a whole pile of complicated expansion conditions, and the
3728 string was silently truncated, leading to an expansion error. It turns out
3729 that it is easier to change to an unlimited length (owing to other changes
3730 that have happened since this code was originally written) than to build
3731 structure for giving a limitation error. The length of the item that
3732 expands into the list of hosts is now unlimited.
3734 47. The lsearch lookup could not handle data where the length of text line was
3735 more than 4095 characters. Such lines were truncated, leading to shortened
3736 data being returned. It should now handle lines of any length.
3738 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3739 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3742 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3743 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3744 no longer happens in comment lines. A list of replacements is now placed
3745 at the head of all of the source files, except those whose only change is
3746 to replace PERL_COMMAND in the very first #! line.
3748 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3749 messages on the queue, with a bottom-up merge sort, using code contributed
3750 by Michael Haardt. This should make operations like -bp somewhat faster on
3751 large queues. It won't affect queue runners, except when queue_run_in_order
3754 51. Installed eximstats 1.31 in the distribution.
3756 52. Added support for SRV lookups to the dnslookup router.
3758 53. If an ACL referred to $message_body or $message_body_end, the value was not
3759 reset for any messages that followed in the same SMTP session.
3761 54. The store-handling optimization for building very long strings was not
3762 differentiating between the different store pools. I don't think this
3763 actually made any difference in practice, but I've tidied it.
3765 55. While running the routers to verify a sender address, $sender_address
3766 was still set to the sender address. This is wrong, because when routing to
3767 send a bounce to the sender, it would be empty. Therefore, I have changed
3768 it so that, while verifying a sender address, $sender_address is set to <>.
3769 (There is no change to what happens when verifying a recipient address.)
3771 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3772 target A or AAAA records (if not already returned) without resetting the
3773 qualify_single or search_parents options of the DNS resolver. These are
3774 inappropriate in this case because the targets of MX and SRV records must
3775 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3776 target that, when qualified, matched something in the local domain. These
3777 two options are now turned off when doing these lookups.
3779 57. It seems that at least some releases of Reiserfs (which does not have the
3780 concept of a fixed number of inodes) returns zero and not -1 for the
3781 number of available inodes. This interacted badly with check_spool_inodes,
3782 which assumed that -1 was the "no such thing" setting. What I have done is
3783 to check that the total number of inodes is greater than zero before doing
3784 the test of how many are available.
3786 58. When a "warn" ACL statement has a log_message modifier, the message is
3787 remembered, and not repeated. This is to avoid a lot of repetition when a
3788 message has many recipients that cause the same warning to be written.
3789 Howewer, Exim was preserving the list of already written lines for an
3790 entire SMTP session, which doesn't seem right. The memory is now reset if a
3791 new message is started.
3793 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3794 debugging output unless log_rewrite was also set.
3796 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3797 of (int)(handle) when we know that handle contains (void *)(-1).
3799 61. The Exim daemon panic-logs an error return when it closes the incoming
3800 connection. However "connection reset by peer" seems to be common, and
3801 isn't really an error worthy of noting specially, so that particular error
3804 62. When Exim is trying to find all the local interfaces, it used to panic and
3805 die if the ioctl to get the interface flags failed. However, it seems that
3806 on at least one OS (Solaris 9) it is possible to have an interface that is
3807 included in the list of interfaces, but for which you get a failure error
3808 for this call. This happens when the interface is not "plumbed" into a
3809 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3810 failure of the "get flags" call assumes that the interface is down.
3812 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3813 makes life easier for people who are doing arithmetic on fields extracted
3814 from dates, where you often get leading zeros that should not be
3815 interpreted as octal.
3817 64. Added qualify_domain to the redirect router, to override the global
3820 65. If a pathologically long header line contained very many addresses (the
3821 report of this problem mentioned 10 000) and each of them was rewritten,
3822 Exim could use up a very large amount of memory. (It kept on making new
3823 copies of the header line as it rewrote, and never released the old ones.)
3824 At the expense of a bit more processing, the header rewriting function has
3825 been changed so that it no longer eats memory in this way.
3827 66. The generation of the Received: header has been moved from the time that a
3828 message starts to be received, to the time that it finishes. The timestamp
3829 in the Received: header should now be very close to that of the <= log
3830 line. There are two side-effects of this change:
3832 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3833 logged header lines no longer include the local Received: line, because
3834 it has not yet been created. The same applies to a copy of the message
3835 that is returned to a non-SMTP sender when a message is rejected.
3837 (b) When a filter file is tested using -bf, no additional Received: header
3838 is added to the test message. After some thought, I decided that this
3841 This change does not affect the value of $received_for. It is still set
3842 after address rewriting, but before local_scan() is called.
3844 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3846 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3847 gave an unhelpful panic error message, and a defer error. I have managed to
3848 change this behaviour so that it now rejects any supplied certificate,
3849 which seems right, as the list of acceptable certificates is empty.
3851 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3852 gave an unhelpful defer error. I have not managed to make this reject any
3853 supplied certificates, but the error message it gives is "no certificate
3854 supplied", which is not helpful.
3856 70. exigrep's output now also includes lines that are not associated with any
3857 message, but which match the given pattern. Implemented by a patch from
3858 Martin Sluka, which also tidied up the Perl a bit.
3860 71. Recipient callout verification, like sender verification, was using <> in
3861 the MAIL FROM command. This isn't really the right thing, since the actual
3862 sender may affect whether the remote host accepts the recipient or not. I
3863 have changed it to use the actual sender in the callout; this means that
3864 the cache record is now keyed on a recipient/sender pair, not just the
3865 recipient address. There doesn't seem to be a real danger of callout loops,
3866 since a callout by the remote host to check the sender would use <>.
3867 [SEE ABOVE: changed after hitting problems.]
3869 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3870 temporary errors. However, in the case of such a code being given after
3871 the end of a data transmission (i.e. after ".") Exim was failing to write
3872 a retry record for the message. (Yes, there was some broken host that was
3873 actually sending 8xx at this point.)
3875 73. An unknown lookup type in a host list could cause Exim to panic-die when
3876 the list was checked. (An example that provoked this was putting <; in the
3877 middle of a list instead of at the start.) If this happened during a DATA
3878 ACL check, a -D file could be left lying around. This kind of configuration
3879 error no longer causes Exim to die; instead it causes a defer errror. The
3880 incident is still logged to the main and panic logs.
3882 74. Buglet left over from Exim 3 conversion. The message "too many messages
3883 in one connection" was written to the rejectlog but not the mainlog, except
3884 when address rewriting (yes!) was being logged.
3886 75. Added write_rejectlog option.
3888 76. When a system filter was run not as root (that is, when system_filter_user
3889 was set), the values of the $n variables were not being returned to the
3890 main process; thus, they were not subsequently available in the $sn
3893 77. Added +return_path_on_delivery log selector.
3895 78. A connection timeout was being treated differently from recipients deferred
3896 when testing hosts_max_try with a message that was older than the host's
3897 retry timeout. (The host should not be counted, thus allowing all hosts to
3898 be tried at least once before bouncing.) This may have been the cause of an
3899 occasionally reported bug whereby a message would remain on the queue
3900 longer than the retry timeout, but would be bounced if a delivery was
3901 forced. I say "may" because I never totally pinned down the problem;
3902 setting up timeout/retry tests is difficult. See also the next item.
3904 79. The ultimate address timeout was not being applied to errors that involved
3905 a combination of host plus message (for example, a timeout on a MAIL
3906 command). When an address resolved to a number of possible hosts, and they
3907 were not all tried for each delivery (e.g. because of hosts_max_try), a
3908 message could remain on the queue longer than the retry timeout.
3910 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3913 81. Fixed an obscure SMTP outgoing bug which required at least the following
3914 conditions: (a) there was another message waiting for the same server;
3915 (b) the server returned 5xx to all RCPT commands in the first message so
3916 that the message was not completed; (c) the server dropped the connection
3917 or gave a negative response to the RSET that Exim sends to abort the
3918 transaction. The observed case was a dropped connection after DATA that had
3919 been sent in pipelining mode. That is, the server had advertised PIPELINING
3920 but was not implementing it correctly. The effect of the bug was incorrect
3921 behaviour, such as trying another host, and this could lead to a crash.
3927 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3928 and daemon.c were passed as pointers to ints; they should have been
3929 pointers to socklen_t variables (which are typically unsigned ints).
3931 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3934 3. Fixed a really odd bug that affected only the testing scheme; patching a
3935 certain fixed string in the binary changed the value of another string that
3936 happened to be identical to the end of the original first string.
3938 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3939 name", it returns that address as the IP address. On some operating
3940 systems (e.g. Solaris), it also passes back the IP address string as the
3941 "host name". However, on others (e.g. Linux), it passes back an empty
3942 string. Exim wasn't checking for this, and was changing the host name to an
3943 empty string, assuming it had been canonicized.
3945 5. Although rare, it is permitted to have more than one PTR record for a given
3946 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3947 all the names associated with an address, because they do in Solaris.
3948 However, it seems that they do not in Linux for data that comes from the
3949 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3950 I found this out when I moved to a new Linux workstation and tried to run
3951 the Exim test suite.
3953 To get round this problem I have changed the code so that it now does its
3954 own call to the DNS to look up PTR records when searching for a host name.
3955 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3956 addresses that are only in /etc/hosts are still found.
3958 This behaviour is, however, controlled by an option called host_lookup_
3959 order, which defaults to "bydns:byaddr". If people want to use the other
3960 order, or indeed, just use one or the other means of lookup, they can
3961 specify it in this variable.
3963 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3964 some operating systems, this comes back from gethostbyaddr() as an empty
3965 string, and this is what Exim used to test for. However, it seems that in
3966 other systems, "." is yielded. Exim now tests for this case too.
3968 7. The values of check_spool_space and check_log_space are now held internally
3969 as a number of kilobytes instead of an absolute number of bytes. If a
3970 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3971 kilobyte. This means that much larger values can be stored.
3973 8. Exim monitor: an attempt to get the action menu when not actually pointing
3974 at a message produces an empty menu entitled "No message selected". This
3975 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3976 no entries in it ("Shell widget menu has zero width and/or height"). So I
3977 have added a single, blank menu entry in this case.
3979 9. Added ${quote_local_part.
3981 10. MIME decoding is now applied to the contents of Subject: header lines when
3984 11. Now that a reference to $sender_host_address automatically causes a reverse
3985 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3986 host lookup before query-style lookups in lists that might use this
3987 variable. This has therefore been abolished, and the "net-" prefix is no
3988 longer necessary for query-style lookups.
3990 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3991 have been a typo for LFLAGS, so it has been changed.
3993 13. The install script calls Exim with "-C /dev/null" in order to find the
3994 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3995 to be output. Howeve, since Exim outputs its version number before the
3996 error, it didn't break the script. It just looked ugly. I fixed this by
3997 always allowing "-C /dev/null" if the caller is root.
3999 14. Ignore overlarge ACL variable number when reading spool file - insurance
4000 against a later release with more variables having written the file.
4002 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4003 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4004 this, as well as the form without the "IPv6" on the front.
4006 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4007 OS/Makefile-Darwin file.
4009 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4010 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4012 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4013 variable. However, some OS use ints for inodes. I've added cast to long int
4014 to get rid of the compiler warning.
4016 19. I had forgotten to lock out "/../" in configuration file names when
4017 ALT_CONFIG_PREFIX was set.
4019 20. Routers used for verification do not need to specify transports. However,
4020 if such a router generated a host list, and callout was configured, Exim
4021 crashed, because it could not find a port number from the (non-existent)
4022 transport. It now assumes port 25 in this circumstance.
4024 21. Added the -t option to exigrep.
4026 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4027 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4028 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4029 them is set, LOOKUP_LSEARCH is forced.
4031 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4033 24. Added sender and host information to the "rejected by local_scan()" log
4034 line; previously there was no indication of these.
4036 25. Added .include_if_exists.
4038 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4039 Linux. It turns out that not all file systems support this. Apparently some
4040 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4041 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4042 not supported on the file descriptor, is now ignored when Exim is trying to
4043 sync a directory. This applies only to Linux.
4045 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4047 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4048 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4051 29. Added support for maildirsize files from supplied patch (modified a bit).
4053 30. The use of :fail: followed by an empty string could lead Exim to respond to
4054 sender verification failures with (e.g.):
4056 550 Verification failed for <xxx>
4057 550 Sender verify failed
4059 where the first response line was missing the '-' that indicates it is not
4060 the final line of the response.
4062 31. The loop for finding the name of the user that called Exim had a hardwired
4063 limit of 10; it now uses the value of finduser_retries, which is used for
4064 all other user lookups.
4066 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4069 33. Exim was neglecting to zero errno before one call of strtol() when
4070 expanding a string and expecting an integer value. On some systems this
4071 resulted in spurious "integer overflow" errors. Also, it was casting the
4072 result into an int without checking.
4074 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4075 did not work. The code looks as if it has *never* worked, though it appears
4076 to have been documented since at least releast 1.62. I have made it work.
4078 35. The "timeout_DNS" error in retry rules, also documented since at least
4079 1.62, also never worked. As it isn't clear exactly what this means, and
4080 clearly it isn't a major issue, I have abolished the feature by treating it
4081 as "timeout", and writing a warning to the main and panic logs.
4083 36. The display of retry rules for -brt wasn't always showing the error code
4086 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4087 timeout_connect_A, timeout_connect_MX.
4089 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4090 to the empty sender.
4092 39. The daemon was not analysing the content of -oX till after it had closed
4093 stderr and disconnected from the controlling terminal. This meant that any
4094 syntax errors were only noted on the panic log, and the return code from
4095 the command was 0. By re-arranging the code a little, I've made the
4096 decoding happen first, so such errors now appear on stderr, and the return
4097 code is 1. However, the actual setting up of the sockets still happens in
4098 the disconnected process, so errors there are still only recorded on the
4101 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4102 connections (as happens on some IP stacks) was logged at start up time as
4103 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4104 it from "IPv6 and IPv4", which means that two separate sockets are being
4107 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4108 says whether AF_INET or AF_INET6 was passed as an argument.
4110 42. Exiwhat output was messed up when time zones were included in log
4113 43. Exiwhat now gives more information about the daemon's listening ports,
4114 and whether -tls-on-connect was used.
4116 44. The "port" option of the smtp transport is now expanded.
4118 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4119 silently ignored. Now an error message is written to the main and panic
4122 46. There's a new ACL modifier called "logwrite" which writes to a log file
4123 as soon as it is encountered.
4125 47. Added $local_user_uid and $local_user_gid at routing time.
4127 48. Exim crashed when trying to verify a sender address that was being
4130 49. Exim was recognizing only a space character after ".include". It now also
4131 recognizes a tab character.
4133 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4134 extracting the relevant information from the specification. The man page no
4135 longer contains scrambled data for the -d option, and I've added a section
4136 at the front about calling Exim under different names.
4138 51. Added "extra_headers" argument to the "mail" command in filter files.
4140 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4143 53. Installed eximstats 1.29.
4145 54. Added transport_filter_timeout as a generic transport option.
4147 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4148 Cc: header lines. This was required by RFC 822, but it not required by RFC
4151 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4152 incoming messages that did not have them. Now it does so only if the
4153 message originates locally, that is, if there is no associated remote host
4154 address. When Resent- header lines are present, this applies to the Resent-
4155 lines rather than the non-Resent- lines.
4157 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4158 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4160 58. Messages for configuration errors now include the name of the main
4161 configuration file - useful now that there may be more than one file in a
4162 list (.included file names were always shown).
4164 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4165 for those rare installations that do not start the daemon as root or run it
4166 setuid root. I've cut out the call to initgroups() if the daemon is not
4169 60. The Exim user and group can now be bound into the binary as text strings
4170 that are looked up at the start of Exim's processing.
4172 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4174 62. Added $mailstore_basename variable.
4176 63. Installed patch to sieve.c from Michael Haardt.
4178 64. When Exim failed to open the panic log after failing to open the main log,
4179 the original message it was trying to log was written to stderr and debug
4180 output, but if they were not available (the usual case in production), it
4181 was lost. Now it is written to syslog before the two lines that record the
4182 failures to open the logs.
4184 65. Users' Exim filters run in subprocesses under the user's uid. It is
4185 possible for a "deliver" command or an alias in a "personal" command to
4186 provoke an address rewrite. If logging of address rewriting is configured,
4187 this fails because the process is not running as root or exim. There may be
4188 a better way of dealing with this, but for the moment (because 4.30 needs
4189 to be released), I have disabled address rewrite logging when running a
4190 filter in a non-root, non-exim process.
4196 1. The buildconfig auxiliary program wasn't quoting the value set for
4197 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4198 not defined. This bug was masked in 4.22 by the effect that was fixed in
4201 2. Some messages that were rejected after a message id was allocated were
4202 shown as "incomplete" by exigrep. It no longer does this for messages that
4203 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4205 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4206 have allow_domain_literals set, the ID did not get logged in the <= line.
4207 Domain literals are now always recognized in Message-ID: header lines.
4209 4. The first argument for a ${extract expansion item is the key name or field
4210 number. Leading and trailing spaces in this item were not being ignored,
4211 causing some misleading effects.
4213 5. When deliver_drop_privilege was set, single queue runner processes started
4214 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4215 same command in the process it spins off) were not dropping privilege.
4217 6. When the daemon running as "exim" started a queue runner, it always
4218 re-executed Exim in the spun-off process. This is a waste of effort when
4219 deliver_drop_privilege is set. The new process now just calls the
4220 queue-runner function directly.
4226 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4229 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4230 ignored. Though the use of -oP was forcing the writing of a pid file, it
4231 was always written to the default place.
4233 3. If the message "no IP address found for host xxxx" is generated during
4234 incoming verification, it is now followed by identification of the incoming
4235 connection (so you can more easily find what provoked it).
4237 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4239 5. Added some features to "harden" Exim a bit more against certain attacks:
4241 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4242 be put in Local/Makefile. This is like the never_users runtime option,
4243 but it cannot be overridden. The default setting is "root".
4245 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4246 prefix string with which any file named in a -C command line option
4249 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4250 is retained for -C and -D only if the caller of Exim is root. Without
4251 it, the exim user may also use -C and -D and retain privilege.
4253 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4254 command line option is disabled.
4256 6. Macro names set by the -D option must start with an upper case letter, just
4257 like macro names defined in the configuration file.
4259 7. Added "dereference=" facility to LDAP.
4261 8. Two instances of the typo "uknown" in the source files are fixed.
4263 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4264 the Configure-Makefile script screwed up while processing it.
4266 10. Incorporated PCRE 4.4.
4268 11. The SMTP synchronization check was not operating right at the start of an
4269 SMTP session. For example, it could not catch a HELO sent before the client
4270 waited for the greeting. There is now a check for outstanding input at the
4271 point when the greeting is written. Because of the duplex, asynchronous
4272 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4273 way, but not yet received, when the check is performed.
4275 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4276 on TCP/IP sockets, because this apparently causes some broken clients to
4279 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4280 unchanged) from the Cygwin maintainer.
4282 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4283 was supported instead of testing for mbx. Effectively a typo.
4285 15. The spa authenticator server code was not checking that the input it
4286 received was valid base64.
4288 16. The debug output line for the "set" modifier in ACLs was not showing the
4289 name of the variable that was being set.
4291 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4293 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4294 lookup on its own. Something else was needed to trigger the lookup. For
4295 example, a match in host_lookup or the need for a host name in a host list.
4296 Now, if $sender_host_name is referenced and the host name has not yet been
4297 looked up, a lookup is performed. If the lookup fails, the variable remains
4298 empty, and $host_lookup_failed is set to "1".
4300 19. Added "eqi" as a case-independent comparison operator.
4302 20. The saslauthd authentication condition could segfault if neither service
4303 nor realm was specified.
4305 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4306 error message that was logged was misleading, and incoming SMTP
4307 connections were dropped. The message is now more accurate, and temporary
4308 errors are given to SMTP connections.
4310 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4311 (except RSET) with a 421 error, until QUIT is received. However, it was
4312 failing to send a response to QUIT.
4314 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4315 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4316 ACL is now run after the helo_try_verify_hosts code.
4318 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4319 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4320 case-independent, so other case variants are also recognized). Apparently
4321 some systems use these upper case variants.
4323 25. If more than two messages were waiting for the same host, and a transport
4324 filter was specified for the transport, Exim sent two messages over the
4325 same TCP/IP connection, and then failed with "socket operation on non-
4326 socket" when it tried to send the third.
4328 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4330 27. The extern definition of crypt16() in expand.c was not being excluded when
4331 the OS had its own crypt16() function.
4333 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4334 as a preferred synonym for return_size_limit, both as an option and as an
4337 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4339 30. Changed the default configuration ACL to relax the local part checking rule
4340 for addresses that are not in any local domains. For these addresses,
4341 slashes and pipe symbols are allowed within local parts, but the sequence
4342 /../ is explicitly forbidden.
4344 31. SPA server authentication was not clearing the challenge buffer before
4347 32. log_message in a "warn" ACL statement was writing to the reject log as
4348 well as to the main log, which contradicts the documentation and doesn't
4349 seem right (because no rejection is happening). So I have stopped it.
4351 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4352 However, I am unable to do any testing of this.
4354 34. Fixed an infelicity in the appendfile transport. When checking directories
4355 for a mailbox, to see if any needed to be created, it was accidentally
4356 using path names with one or more superfluous leading slashes; tracing
4357 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4359 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4360 contents are added to the log line that is written for every discarded
4361 recipient. (Previously a log_message setting was ignored.)
4363 36. The ${quote: operator now quotes the string if it is empty.
4365 37. The install script runs exim in order to find its version number. If for
4366 some reason other than non-existence or emptiness, which it checks, it
4367 could not run './exim', it was installing it with an empty version number,
4368 i.e. as "exim-". This error state is now caught, and the installation is
4371 38. An argument was missing from the function that creates an error message
4372 when Exim fails to connect to the socket for saslauthd authentication.
4373 This could cause Exim to crash, or give a corrupted message.
4375 39. Added isip, isip4, and isip6 to ${if conditions.
4377 40. The ACL variables $acl_xx are now saved with the message, and can be
4378 accessed later in routers, transports, and filters.
4380 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4381 strings in the file are not string-expanded.
4383 42. If a MAIL command specified a SIZE value that was too large to fit into an
4384 int variable, the check against message_size_limit failed. Such values are
4385 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4386 day this will have to be increased, but I don't think I want to be around
4387 when emails are that large.
4394 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4395 iconv() is not standard in FreeBSD.
4397 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4398 IPv6 enabled. The observed symptom was a segmentation fault on return from
4399 the function os_common_find_running_interfaces() in src/os.c.
4401 3. In the check_special_case() function in daemon.c I had used "errno" as an
4402 argument name, which causes warnings on some systems. This was basically a
4403 typo, since it was named "eno" in the comments!
4405 4. The code that waits for the clock to tick (at a resolution of some fraction
4406 of a second) so as to ensure message-id uniqueness was always waiting for
4407 at least one whole tick, when it could have waited for less. [This is
4408 almost certainly not relevant at current processor speeds, where it is
4409 unlikely to ever wait at all. But we try to future-proof.]
4411 5. The function that sleeps for a time interval that includes fractions of a
4412 second contained a race. It did not block SIGALRM between setting the
4413 timer, and suspending (a couple of lines later). If the interval was short
4414 and the sigsuspend() was delayed until after it had expired, the suspension
4415 never ended. On busy systems this could lead to processes getting stuck for
4418 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4419 process, before it forks any delivery processes. The open lookup caching
4420 mechanism meant that the open file or database connection was passed into
4421 the delivery process. The problem was that delivery processes always tidy
4422 up cached lookup data. This could cause a problem for the next delivery
4423 process started by the queue runner, because the external queue runner
4424 process does not know about the closure. So the next delivery process
4425 still has data in the lookup cache. In the case of a file lookup, there was
4426 no problem because closing a file descriptor in a subprocess doesn't affect
4427 the parent. However, if the lookup was caching a connection to a database,
4428 the connection was closed, and the second delivery process was likely to
4429 see errors such as "PGSQL: query failed: server closed the connection
4430 unexpectedly". The problem has been fixed by closing all cached lookups
4431 in a queue runner before running a delivery process.
4433 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4434 seem to have the "const" qualifier which it has on other OS. I've
4437 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4438 *of the same type* (client or server) with the same public name that an
4439 error should be diagnosed.
4441 9. When Exim looked up a host name for an IP address, but failed to find the
4442 original IP address when looking up the host name (a safety check), it
4443 output the message "<ip address> does not match any IP for NULL", which was
4444 confusing, to say the least. The bug was that the host name should have
4445 appeared instead of "NULL".
4447 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4448 user that is built into the binary, and the -C or -D options is used, root
4449 privilege is dropped before the configuration file is read. In addition,
4450 logging is switched to stderr instead of the normal log files. If the
4451 configuration then re-defines the Exim user, the unprivileged environment
4452 is probably not what is expected, so Exim logs a panic warning message (but
4455 However, if deliver_drop_privilege is set, the unprivileged state may well
4456 be exactly what is intended, so the warning has been cut out in that case,
4457 and Exim is allowed to try to write to its normal log files.
4463 1. smtp_return_error_details was not giving details for temporary sender
4464 or receiver verification errors.
4466 2. Diagnose a configuration error if two authenticators have the same public
4469 3. Exim used not to create the message log file for a message until the first
4470 delivery attempt. This could be confusing when incoming messages were held
4471 for policy or load reasons. The message log file is now created at the time
4472 the message is received, and an initial "Received" line is written to it.
4474 4. The automatically generated man page for command line options had a minor
4475 bug that caused no ill effects; however, a more serious problem was that
4476 the procedure for building the man page automatically didn't always
4477 operate. Consequently, release 4.20 contains an out-of-date version. This
4478 shouldn't happen again.
4480 5. When building Exim with embedded Perl support, the script that builds the
4481 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4482 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4484 6. The freeze_tell option was not being used for messages that were frozen on
4485 arrival, either by an ACL or by local_scan().
4487 7. Added the smtp_incomplete_transaction log selector.
4489 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4490 was accepting AUTH without a new EHLO.
4492 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4493 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4496 10. Logging of TCP/IP connections (when configured) now happens in the main
4497 daemon process instead of the child process, so that the TCP/IP connection
4498 count is more accurate (but it can never be perfect).
4500 11. The use of "drop" in a nested ACL was not being handled correctly in the
4501 outer ACL. Now, if condition failure induced by the nested "drop" causes
4502 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4503 or "require"), the connection is dropped.
4505 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4506 that yield "discard" can now be used with an "accept" or a "discard" verb,
4507 but an error is generated for any others (because I can't see a useful way
4508 to define what should happen).
4510 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4511 are now processed in the same way as lines in the Exim configuration file.
4512 In particular, continuation lines are supported.
4514 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4516 15. Added -ti meaning -t -i.
4518 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4519 domains, and warn by logging if others are found.
4521 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4522 returned from GIFCONF when Exim is trying to find the list of interfaces on
4523 a host. The code in os.c has been modified to copy each ifreq to an aligned
4524 structure in all cases.
4526 Also, in some cases, the returned ifreq's were being copied to a 'struct
4527 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4528 means the last couple of bytes of an IPv6 address could be chopped if the
4529 ifreq contained only a normal sockaddr (14 bytes storage).
4531 18. Named domain lists were not supported in the hosts_treat_as_local option.
4532 An entry such as +xxxx was not recognized, and was treated as a literal
4535 19. Ensure that header lines added by a DATA ACL are included in the reject log
4536 if the ACL subsequently rejects the message.
4538 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4539 MD5 (which is deprecated).
4541 21. When testing a filter file using -bf, Exim was writing a message when it
4542 took the sender from a "From " line in the message, but it was not doing so
4543 when it took $return_path from a Return-Path: header line. It now does.
4545 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4546 with a valid header line field name (a series of printing characters
4547 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4549 23. Changed "disc" in the source to "disk" to conform to the documentation and
4550 the book and for uniformity.
4552 24. Ignore Sendmail's -Ooption=value command line item.
4554 25. When execve() failed while trying to run a command in a pipe transport,
4555 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4556 could be confused with a return value of 69 from the command itself. This
4557 has been changed to 127, the value the shell returns if it is asked to run
4558 a non-existent command. The wording for the related log line suggests a
4559 non-existent command as the problem.
4561 26. If received_header_text expands to an empty string, do not add a Received:
4562 header line to the message. (Well, it adds a token one on the spool, but
4563 marks it "old" so that it doesn't get used or transmitted.)
4565 27. Installed eximstats 1.28 (addition of -nt option).
4567 28. There was no check for failure on the call to getsockname() in the daemon
4568 code. This can fail if there is a shortage of resources on the system, with
4569 ENOMEM, for example. A temporary error is now given on failure.
4571 29. Contrary to the C standard, it seems that in some environments, the
4572 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4573 program. Exim now does this explicitly; it affects the formatting of
4574 timestamps using strftime().
4576 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4577 complaints. I've now initialized all the variables, to avoid this.
4579 32. Header lines added by a system filter were not being "seen" during
4580 transport-time rewrites.
4582 33. The info_callback() function passed to OpenSSL is set up with type void
4583 (*)(SSL *, int, int), as described somewhere. However, when calling the
4584 function (actually a macro) that sets it up, the type void(*)() is
4585 expected. I've put in a cast to prevent warnings from picky compilers.
4587 34. If a DNS black list lookup found a CNAME record, but there were no A
4588 records associated with the domain it pointed at, Exim crashed.
4590 35. If a DNS black list lookup returned more than one A record, Exim ignored
4591 all but the first. It now scans all returned addresses if a particular IP
4592 value is being sought. In this situation, the contents of the
4593 $dnslist_value variable are a list of all the addresses, separated by a
4596 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4597 to accept a host name and all its aliases if the forward lookup for any of
4598 them yielded the IP address of the incoming connection. Now it accepts only
4599 those names whose forward lookup yields the correct IP address. Any other
4600 names are discarded. This closes a loophole whereby a rogue DNS
4601 administrator could create reverse DNS records to break through a
4602 wildcarded host restriction in an ACL.
4604 37. If a user filter or a system filter that ran in a subprocess used any of
4605 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4606 the wrong values were passed to the pipe command ($thisaddress had the
4607 value of $0, $0 had the value of $1, etc). This bug was introduced by
4608 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4610 38. Improved the line breaking for long SMTP error messages from ACLs.
4611 Previously, if there was no break point between 40 and 75 characters, Exim
4612 left the rest of the message alone. Two changes have been made: (a) I've
4613 reduced the minimum length to 35 characters; (b) if it can't find a break
4614 point between 35 and 75 characters, it looks ahead and uses the first one
4615 that it finds. This may give the occasional overlong line, but at least the
4616 remaining text gets split now.
4618 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4619 file descriptors might be low, and that setting 1000 would always raise it.
4620 It turns out that in some environments, the limit is already over 1000 and
4621 that lowering it causes trouble. So now Exim takes care not to decrease it.
4623 40. When delivering a message, the value of $return_path is set to $sender_
4624 address at the start of routing (routers may change the value). By an
4625 oversight, this default was not being set up when an address was tested by
4626 -bt or -bv, which affected the outcome if any router or filter referred to
4629 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4630 log only when "message" or "log_message" are set. However, if one of the
4631 conditions was an address verification, or a call to a nested ACL, the
4632 messages generated by the underlying test were being passed through. This
4633 no longer happens. The underlying message is available in $acl_verify_
4634 message for both "message" and "log_message" expansions, so it can be
4635 passed through if needed.
4637 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4638 new expansion $bh_ to give the encoded byte string without charset
4639 translation. Translation happens only if iconv() is available; HAVE_ICONV
4640 indicates this at build time. HEADERS_CHARSET gives the charset to
4641 translate to; headers_charset can change it in the configuration, and
4642 "headers charset" can change it in an individual filter file.
4644 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4645 that creates RFC 2047 encoded "words" labels them as that charset instead
4646 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4647 expansion operator; (ii) when Exim creates a From: line for a local
4648 message; (iii) when a header line is rewritten to include a "phrase" part.
4650 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4651 buggy, causing it to skip the first lines of messages whose message ID
4652 ended in 'D'. This would not have bitten before Exim release 4.14, because
4653 message IDs were unlikely to end in 'D' before then. The effect was to have
4654 incorrect size information for certain domains.
4656 45. #include "config.h" was missing at the start of the crypt16.c module. This
4657 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4660 46. If there was a timeout during a "random" callout check, Exim treated it as
4661 a failure of the random address, and carried on sending RSET and the real
4662 address. If the delay was just some slowness somewhere, the response to the
4663 original RCPT would be taken as a response to RSET and so on, causing
4664 mayhem of various kinds.
4666 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4667 when I implemented it. It didn't allow for the fact that some option values
4668 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4669 the right test for positive values.
4671 48. Domain names in DNS records are case-independent. Exim always looks them up
4672 in lower case. Some resolvers return domain names in exactly the case they
4673 appear in the zone file, that is, they may contain uppercase letters. Not
4674 all resolvers do this - some return always lower case. Exim was treating a
4675 change of case by a resolver as a change of domain, similar to a widening
4676 of a domain abbreviation. This triggered its re-routing code and so it was
4677 trying to route what was effectively the same domain again. This normally
4678 caused routing to fail (because the router wouldn't handle the domain
4679 twice). Now Exim checks for this case specially, and just changes the
4680 casing of the domain that it ultimately uses when it transmits the message
4683 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4686 50. If a filter generated a file delivery with a non-absolute name (possible if
4687 no home directory exists for the router), the forbid_file option was not
4690 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4691 the existing equality matching.
4693 52. Exim was using ints instead of ino_t variables in some places where it was
4694 dealing with inode numbers.
4696 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4697 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4698 called TMPDIR, and if it finds it is different, it changes its value.
4700 54. The smtp_printf() function is now made available to local_scan() so
4701 additional output lines can be written before returning. There is also an
4702 smtp_fflush() function to enable the detection of a dropped connection.
4703 The variables smtp_input and smtp_batched_input are exported to
4706 55. Changed the default runtime configuration: the message "Unknown user"
4707 has been removed from the ACL, and instead placed on the localuser router,
4708 using the cannot_route_message feature. This means that any verification
4709 failures that generate their own messages won't get overridden. Similarly,
4710 the "Unrouteable address" message that was in the ACL for unverifiable
4711 relay addresses has also been removed.
4713 56. Added hosts_avoid_esmtp to the smtp transport.
4715 57. The exicyclog script was not checking for the esoteric option
4716 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4717 will work only if exicyclog is run under the appropriate euid.
4719 58. Following a discussion on the list, the rules by which Exim recognises line
4720 endings on incoming messages have been changed. The -dropcr and drop_cr
4721 options are now no-ops, retained only for backwards compatibility. The
4722 following line terminators are recognized: LF CRLF CR. However, special
4723 processing applies to CR:
4725 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4726 nor a local message in the state where . is a terminator.
4728 (ii) If a bare CR is encountered in a header line, an extra space is added
4729 after the line terminator so as not to end the header. The reasoning
4730 behind this is that bare CRs in header lines are most likely either
4731 to be mistakes, or people trying to play silly games.
4733 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4734 was being incorrectly given as 18 bytes larger than it should have been.
4735 This is a VOB (very old bug).
4737 60. This may never have affected anything current, but just in case it has:
4738 When the local host is found other than at the start of a list of hosts,
4739 the local host, those with the same MX, and any that follow, are discarded.
4740 When the list in question was part of a longer list of hosts, the following
4741 hosts (not currently being processed) were also being discarded. This no
4742 longer happens. I'm not sure if this situation could ever has previously
4745 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4748 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4749 header. This is something that is recommended in a new Internet Draft, and
4750 is something that is documented as being done by Sendmail. There are two
4751 possible values. For messages generated by the autoreply transport, Exim
4754 Auto-Submitted: auto-replied
4756 whereas for all other generated messages (e.g. bounces) it adds
4758 Auto-Submitted: auto-generated
4760 63. The "personal" condition in filters now includes a test for the
4761 Auto-Submitted: header. If it contains the string "auto-" the message it
4762 not considered personal.
4764 64. Added rcpt_include_affixes as a generic transport option.
4766 65. Added queue_only_override (default true).
4768 66. Added the syslog_duplication option.
4770 67. If what should have been the first header line of a message consisted of
4771 a space followed by a colon, Exim was mis-interpreting it as a header line.
4772 It isn't of course - it is syntactically invalid and should therefore be
4773 treated as the start of the message body. The misbehaviour could have
4774 caused a number of strange effects, including loss of data in subsequent
4775 header lines, and spool format errors.
4777 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4778 client host had authenticated. This control can now be exercised by an ACL
4779 for more flexibility.
4781 69. By default, callouts do not happen when testing with -bh. There is now a
4782 variant, -bhc, which does actually run the callout code, including
4783 consulting and updating the callout cache.
4785 70. Added support for saslauthd authentication, courtesy of Alexander
4788 71. If statvfs() failed on the spool or log directories while checking their
4789 size for availability, Exim confusingly gave the error "space shortage".
4790 Furthermore, in debugging mode it crashed with a floating point exception.
4791 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4792 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4793 serious problem, Exim now writes to the main and panic logs when this
4794 happens, with details of the failure. It then refuses to accept the
4795 incoming message, giving the message "spool directory problem" or "log
4796 directory problem" with a 421 code for SMTP messages.
4798 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4799 0, 1, and 2 exist, because some OS complain for execs without them (see
4800 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4801 descriptors. However, the code omitted to check that the open succeeded,
4802 causing mysterious errors if for some reason the permissions on /dev/null
4803 got screwed. Now Exim writes a message to the main and panic logs, and
4804 bombs out if it can't open /dev/null.
4806 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4807 interact so that it is all more flexible. It is supposed to remain
4808 backwards compatible. Also added extra_local_interfaces.
4810 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4811 to bomb out with an assertion failure - to the client this appears as a
4812 connection drop. This problem occurs in the part of the code that was taken
4813 from the Samba project. Fortunately, the assertion is in a very simple
4814 function, so I have fixed this by reproducing the function inline in the
4815 one place where it is called, and arranging for authentication to fail
4816 instead of killing the process with assert().
4818 75. The SPA client code was not working when the server requested OEM rather
4819 than Unicode encoding.
4821 76. Added code to make require_files with a specific uid setting more usable in
4822 the case where statting the file as root fails - usually a non-root-mounted
4823 NFS file system. When this happens and the failure is EACCES, Exim now
4824 forks a subprocess and does the per-uid checking as the relevant uid.
4826 77. Added process_log_path.
4828 78. If log_file_path was not explicitly set, a setting of check_log_space or
4829 check_log_inodes was ignored.
4831 79. If a space check for the spool or log partitions fails, the incident is now
4832 logged. Of course, in the latter case the data may get lost...
4834 80. Added the %p formatting code to string_format() so that it can be used to
4835 print addresses in debug_print(). Adjusted all the address printing in the
4836 debugging in store.c to use %p rather than %d.
4838 81. There was a concern that a line of code in smtp_in.c could overflow a
4839 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4840 initially expressed, the concern was not well-founded, because trailing
4841 spaces are removed early. However, if the trailing spaces were followed by
4842 a NULL, they did not get removed, so the overflow was possible. Two fixes
4845 (a) I re-wrote the offending code in a cleaner fashion.
4846 (b) If an incoming SMTP command contains a NULL character, it is rejected
4849 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4850 runs initgroups(), so that if the Exim user is in any additional groups,
4851 they will be used during message reception.
4857 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob