1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.544 2008/02/12 12:52:51 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
10 When building exim an external PCRE library is now needed -
11 PCRE is a system library on the majority of modern systems.
12 See entry on PCRE_LIBS in EDITME file.
14 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
15 conversation. Added nologin parameter to request.
16 Patch contributed by Kirill Miazine
18 TF/01 Do not log submission mode rewrites if they do not change the address.
20 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
22 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
23 log files in place. Contributed by Roberto Lima
25 NM/04 Bugzilla 667: close socket used by dovecot authenticator
27 TF/03 Bugzilla 615: When checking the local_parts router precondition
28 after a local_part_suffix or local_part_prefix option, Exim now
29 does not use the address's named list lookup cache, since this
30 contains cached lookups for the whole local part.
32 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
33 Robert Millan. Documentation is in experimental-spec.txt
40 TK/01 Add preliminary DKIM support. Currently requires a forked version of
41 ALT-N's libdkim that I have put here:
42 http://duncanthrax.net/exim-experimental/
44 Note to Michael Haardt: I had to rename some vars in sieve.c. They
45 were called 'true' and it seems that C99 defines that as a reserved
46 keyword to be used with 'bool' variable types. That means you could
47 not include C99-style headers which use bools without triggering
48 build errors in sieve.c.
50 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
51 as mailq or other aliases. Changed the --help handling significantly
52 to do whats expected. exim_usage() emits usage/help information.
54 SC/01 Added the -bylocaldomain option to eximstats.
56 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
58 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
60 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
66 PH/01 Another patch from the Sieve maintainer.
68 PH/02 When an IPv6 address is converted to a string for single-key lookup
69 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
70 dots are used instead of colons so that keys in lsearch files need not
71 contain colons. This was done some time before quoting was made available
72 in lsearch files. However, iplsearch files do require colons in IPv6 keys
73 (notated using the quote facility) so as to distinguish them from IPv4
74 keys. This meant that lookups for IP addresses in host lists did not work
75 for iplsearch lookups.
77 This has been fixed by arranging for IPv6 addresses to be expressed with
78 colons if the lookup type is iplsearch. This is not incompatible, because
79 previously such lookups could never work.
81 The situation is now rather anomolous, since one *can* have colons in
82 ordinary lsearch keys. However, making the change in all cases is
83 incompatible and would probably break a number of configurations.
85 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
88 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
89 conversion specification without a maximum field width, thereby enabling
90 a rogue spamd server to cause a buffer overflow. While nobody in their
91 right mind would setup Exim to query an untrusted spamd server, an
92 attacker that gains access to a server running spamd could potentially
93 exploit this vulnerability to run arbitrary code as the Exim user.
95 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
96 $primary_hostname instead of what libspf2 thinks the hosts name is.
98 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
99 a directory entry by the name of the lookup key. Previously, if a
100 symlink pointed to a non-existing file or a file in a directory that
101 Exim lacked permissions to read, a lookup for a key matching that
102 symlink would fail. Now it is enough that a matching directory entry
103 exists, symlink or not. (Bugzilla 503.)
105 PH/03 The body_linecount and body_zerocount variables are now exported in the
108 PH/04 Added the $dnslist_matched variable.
110 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
111 This means they are set thereafter only if the connection becomes
114 PH/06 Added the client_condition to authenticators so that some can be skipped
115 by clients under certain conditions.
117 PH/07 The error message for a badly-placed control=no_multiline_responses left
118 "_responses" off the end of the name.
120 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
122 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
123 (without spaces) instead of just copying the configuration text.
125 PH/10 Added the /noupdate option to the ratelimit ACL condition.
127 PH/11 Added $max_received_linelength.
129 PH/12 Added +ignore_defer and +include_defer to host lists.
131 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
132 way in which PCRE > 7.0 is built.
134 PH/14 Implemented queue_only_load_latch.
136 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
137 MAIL command. The effect was to mangle the value on 64-bit systems.
139 PH/16 Another patch from the Sieve maintainer.
141 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
143 PH/18 If a system quota error occurred while trying to create the file for
144 a maildir delivery, the message "Mailbox is full" was not appended to the
145 bounce if the delivery eventually timed out. Change 4.67/27 below applied
146 only to a quota excession during the actual writing of the file.
148 PH/19 It seems that peer DN values may contain newlines (and other non-printing
149 characters?) which causes problems in log lines. The DN values are now
150 passed through string_printing() before being added to log lines.
152 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
153 and InterBase are left for another time.)
155 PH/21 Added message_body_newlines option.
157 PH/22 Guard against possible overflow in moan_check_errorcopy().
159 PH/23 POSIX allows open() to be a macro; guard against that.
161 PH/24 If the recipient of an error message contained an @ in the local part
162 (suitably quoted, of course), incorrect values were put in $domain and
163 $local_part during the evaluation of errors_copy.
169 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
170 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
173 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
174 issue a MAIL command.
176 PH/02 In an ACL statement such as
178 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
180 if a client was not listed at all, or was listed with a value other than
181 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
182 the condition was not true (as it should be), so access was not denied.
183 The bug was that the ! inversion was incorrectly passed on to the second
184 item. This has been fixed.
186 PH/03 Added additional dnslists conditions == and =& which are different from
187 = and & when the dns lookup returns more than one IP address.
189 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
190 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
192 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
193 FSYNC, which compiles an option called disable_fsync that allows for
194 bypassing fsync(). The documentation is heavily laced with warnings.
196 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
198 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
199 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
200 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
201 including adding "make clean"; (3) Added -fPIC when compiling the test
202 dynamically loaded module, to get rid of a warning.
204 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
205 message fails, move_frozen_messages = true and ignore_bounce_errors_after
206 = 0s. The bug is otherwise harmless.
208 PH/07 There was a bug in the dovecot authenticator such that the value of
209 $auth1 could be overwritten, and so not correctly preserved, after a
210 successful authentication. This usually meant that the value preserved by
211 the server_setid option was incorrect.
213 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
215 PH/09 Installed PCRE release 7.0.
217 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
218 run for batched SMTP input. It is now run at the start of every message
219 in the batch. While fixing this I discovered that the process information
220 (output by running exiwhat) was not always getting set for -bs and -bS
221 input. This is fixed, and it now also says "batched" for BSMTP.
223 PH/11 Added control=no_pipelining.
225 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
226 patch, slightly modified), and move the expansion of helo_data till after
227 the connection is made in the smtp transport (so it can use these
230 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
232 PH/14 Added log_selector = +pid.
234 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
236 PH/16 Add ${if forany and ${if forall.
238 PH/17 Added dsn_from option to vary the From: line in DSNs.
240 PH/18 Flush SMTP output before performing a callout, unless control =
241 no_callout_flush is set.
243 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
244 was true (the default) a successful delivery failed to delete the retry
245 item, thus causing premature timeout of the address. The bug is now
248 PH/20 Added hosts_avoid_pipelining to the smtp transport.
250 PH/21 Long custom messages for fakedefer and fakereject are now split up
251 into multiline reponses in the same way that messages for "deny" and
252 other ACL rejections are.
254 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
255 with slight modification.
257 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
258 draft, changing the syntax and factoring some duplicate code".
260 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
261 for deliveries of the second and subsequent messages over the same SMTP
264 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
265 ${reduce, with only minor "tidies".
267 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
269 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
270 expansion side effects.
272 PH/27 When a message times out after an over-quota error from an Exim-imposed
273 quota, the bounce message says "mailbox is full". This message was not
274 being given when it was a system quota that was exceeded. It now should
277 MH/03 Made $recipients available in local_scan(). local_scan() already has
278 better access to the recipient list through recipients_list[], but
279 $recipients can be useful in postmaster-provided expansion strings.
281 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
282 in the case of a MAIL command with additional options following the
283 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
284 were accidentally chopped off.
286 PH/29 SMTP synchronization checks are implemented when a command is read -
287 there is a check that no more input is waiting when there shouldn't be
288 any. However, for some commands, a delay in an ACL can mean that it is
289 some time before the response is written. In this time, more input might
290 arrive, invalidly. So now there are extra checks after an ACL has run for
291 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
292 pipelining has not been advertised.
294 PH/30 MH's patch to allow iscntrl() characters to be list separators.
296 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
297 returned in the SMTP response when smtp_return_error_details was false.
300 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
301 instead of the C I/O that was originally supplied, because problems were
304 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
305 Exim which did not show up earlier: it was assuming that a call to
306 SSL_CTX_set_info_callback() might give an error value. In fact, there is
307 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
308 was a macro that became an assignment, so it seemed to work. This has
309 changed to a proper function call with a void return, hence the compile
310 error. Exim's code has been fixed.
312 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
315 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
317 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
319 PH/37 If a message is not accepted after it has had an id assigned (e.g.
320 because it turns out to be too big or there is a timeout) there is no
321 "Completed" line in the log. When some messages of this type were
322 selected by exigrep, they were listed as "not completed". Others were
323 picked up by some special patterns. I have improved the selection
324 criteria to be more general.
326 PH/38 The host_find_failed option in the manualroute router can now be set
327 to "ignore", to completely ignore a host whose IP address cannot be
328 found. If all hosts are ignored, the behaviour is controlled by the new
329 host_all_ignored option.
331 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
332 homing or because of multiple MX records with /mx) generated more than
333 one IP address, and the following item turned out to be the local host,
334 all the secondary addresses of the first item were incorrectly removed
335 from the list, along with the local host and any following hosts (which
336 is what is supposed to happen).
338 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
339 whoever called Exim into the -H file. In the case of the daemon it was
340 behaving confusingly. When first started, it used values for whoever
341 started the daemon, but after a SIGHUP it used the Exim user (because it
342 calls itself on a restart). I have changed the code so that it now always
345 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
346 message are rejected with the same error (e.g. no authentication or bad
347 sender address), and a DATA command is nevertheless sent (as can happen
348 with PIPELINING or a stupid MUA), the error message that was given to the
349 RCPT commands is included in the rejection of the DATA command. This is
350 intended to be helpful for MUAs that show only the final error to their
353 PH/42 Another patch from the Sieve maintainer.
355 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
356 Eximstats - Fixed some broken HTML links and added missing column headers
358 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
359 columns for Rejects, Temp Rejects, Ham, and Spam rows.
361 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
363 PH/43 Yet another patch from the Sieve maintainer.
365 PH/44 I found a way to check for a TCP/IP connection going away before sending
366 the response to the final '.' that terminates a message, but only in the
367 case where the client has not sent further data following the '.'
368 (unfortunately, this is allowed). However, in many cases there won't be
369 any further data because there won't be any more messages to send. A call
370 to select() can be used: if it shows that the input is "ready", there is
371 either input waiting, or the socket has been closed. An attempt to read
372 the next input character can distinguish the two cases. Previously, Exim
373 would have sent an OK response which the client would never have see.
374 This could lead to message repetition. This fix should cure that, at
375 least in a lot of common cases.
377 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
378 advertised in response to EHLO.
384 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
385 fixed by 4.65/MH/01 (is this a record?) are fixed:
387 (i) An empty string was always treated as zero by the numeric comparison
388 operators. This behaviour has been restored.
390 (ii) It is documented that the numeric comparison operators always treat
391 their arguments as decimal numbers. This was broken in that numbers
392 starting with 0 were being interpreted as octal.
394 While fixing these problems I realized that there was another issue that
395 hadn't been noticed. Values of message_size_limit (both the global option
396 and the transport option) were treated as octal if they started with 0.
397 The documentation was vague. These values are now always treated as
398 decimal, and I will make that clear in the documentation.
404 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
405 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
408 MH/01 Don't check that the operands of numeric comparison operators are
409 integers when their expansion is in "skipping" mode (fixes bug
410 introduced by 4.64-PH/07).
412 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
413 child addresses, Exim now panics and dies. Previously, because the count
414 is held in a short int, deliveries were likely to be lost. As such a
415 large number of recipients for a single message is ridiculous
416 (performance will be very, very poor), I have chosen to impose a limit
417 rather than extend the field.
423 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
424 leftover -K file (the existence of which was triggered by #402).
425 While we were at it, introduced process PID as part of the -K
426 filename. This should rule out race conditions when creating
429 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
430 processing considerably. Previous code took too long for large mails,
431 triggering a timeout which in turn triggers #401.
433 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
434 in the DK code in transports.c. sendfile() is not really portable,
435 hence the _LINUX specificness.
437 TF/01 In the add_headers option to the mail command in an Exim filter,
438 there was a bug that Exim would claim a syntax error in any
439 header after the first one which had an odd number of characters
442 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
443 callout verification, Exim cached a "reject" for the entire domain. This
444 is correct for most verifications, but it is not correct for a recipient
445 verification with use_sender or use_postmaster set, because in that case
446 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
447 case of MAIL FROM:<> rejection from other early rejections (e.g.
448 rejection of HELO). When verifying a recipient using a non-null MAIL
449 address, the cache is ignored if it shows MAIL FROM:<> rejection.
450 Whatever the result of the callout, the value of the domain cache is
451 left unchanged (for any other kind of callout, getting as far as trying
452 RCPT means that the domain itself is ok).
454 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
457 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
458 manifest itself as EPIPE rather than ECONNECT. When tidying away a
459 session, the daemon ignores ECONNECT errors and logs others; it now
460 ignores EPIPE as well.
462 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
463 (quoted-printable decoding).
465 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
466 later the small subsequent patch to fix an introduced bug.
468 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
470 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
472 PH/08 An error is now given if message_size_limit is specified negative.
474 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
475 to be given (somewhat) arbitrary names.
477 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
480 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
481 miscellaneous code fixes
483 PH/10 Added the log_reject_target ACL modifier to specify where to log
486 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
487 hostname. This is wrong, because it relates to the incoming message (and
488 probably the interface on which it is arriving) and not to the outgoing
489 callout (which could be using a different interface). This has been
490 changed to use the value of the helo_data option from the smtp transport
491 instead - this is what is used when a message is actually being sent. If
492 there is no remote transport (possible with a router that sets up host
493 addresses), $smtp_active_hostname is used.
495 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
496 tweaks were necessary in order to get it to work (see also 21 below):
497 (a) The code assumed that strncpy() returns a negative number on buffer
498 overflow, which isn't the case. Replaced with Exim's string_format()
500 (b) There were several signed/unsigned issues. I just did the minimum
501 hacking in of casts. There is scope for a larger refactoring.
502 (c) The code used strcasecmp() which is not a standard C function.
503 Replaced with Exim's strcmpic() function.
504 (d) The code set only $1; it now sets $auth1 as well.
505 (e) A simple test gave the error "authentication client didn't specify
506 service in request". It would seem that Dovecot has changed its
507 interface. Fortunately there's a specification; I followed it and
508 changed what the client sends and it appears to be working now.
510 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
513 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
514 address A is aliased to B and C, where B exists and C does not. Without
515 -v the output is "A verified" because verification stops after a
516 successful redirection if more than one address is generated. However,
517 with -v the child addresses are also verified. Exim was outputting "A
518 failed to verify" and then showing the successful verification for C,
519 with its parentage. It now outputs "B failed to verify", showing B's
520 parentage before showing the successful verification of C.
522 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
523 look up a TXT record in a specific list after matching in a combined
526 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
527 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
528 they consult the DNS. I had assumed they would set it the way they
529 wanted; and indeed my experiments on Linux seem to show that in some
530 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
531 To be on the safe side, however, I have now made the interface to
532 host_find_byname() similar to host_find_bydns(), with an argument
533 containing the DNS resolver options. The host_find_byname() function now
534 sets these options at its start, just as host_find_bydns() does. The smtp
535 transport options dns_qualify_single and dns_search_parents are passed to
536 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
537 of host_find_byname() use the default settings of RES_DEFNAMES
538 (qualify_single) but not RES_DNSRCH (search_parents).
540 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
541 spool_read_header() do less string testing, by means of a preliminary
542 switch on the second character of optional "-foo" lines. (This is
543 overdue, caused by the large number of possibilities that now exist.
544 Originally there were few.) While I was there, I also converted the
545 str(n)cmp tests so they don't re-test the leading "-" and the first
546 character, in the hope this might squeeze out yet more improvement.
548 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
549 flag allowing group syntax was set by the header_syntax check but not
550 turned off, possible causing trouble later; (2) The flag was not being
551 set at all for the header_verify test, causing "group"-style headers to
552 be rejected. I have now set it in this case, and also caused header_
553 verify to ignore an empty address taken from a group. While doing this, I
554 came across some other cases where the code for allowing group syntax
555 while scanning a header line wasn't quite right (mostly, not resetting
556 the flag correctly in the right place). These bugs could have caused
557 trouble for malformed header lines. I hope it is now all correct.
559 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
560 with the "reply" argument non-NULL. The code, however (which originally
561 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
562 but it didn't always do it. This confused somebody who was copying the
563 code for some other use. I have removed all the tests.
565 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
566 feature that was used to support insecure browsers during the U.S. crypto
567 embargo. It requires special client support, and Exim is probably the
568 only MTA that supported it -- and would never use it because real RSA is
569 always available. This code has been removed, because it had the bad
570 effect of slowing Exim down by computing (never used) parameters for the
571 RSA_EXPORT functionality.
573 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
574 authenticator to fail if there's a tab character in the incoming data
575 (there should never be unless someone is messing about, as it's supposed
576 to be base64-encoded). Also added, on Timo's advice, the "secured" option
577 if the connection is using TLS or if the remote IP is the same as the
578 local IP, and the "valid-client-cert option" if a client certificate has
581 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
582 authenticators. This can be used for authorization after authentication
583 succeeds. (In the case of plaintext, it servers for both authentication
586 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
587 if any retry times were supplied.
589 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
590 connection, where there is no client IP address to check. In this
591 situation, the verify now always succeeds.
593 PH/25 Applied John Jetmore's -Mset patch.
595 PH/26 Added -bem to be like -Mset, but loading a message from a file.
597 PH/27 In a string expansion for a processed (not raw) header when multiple
598 headers of the same name were present, leading whitespace was being
599 removed from all of them, but trailing whitespace was being removed only
600 from the last one. Now trailing whitespace is removed from each header
601 before concatenation. Completely empty headers in a concatenation (as
604 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
605 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
607 PH/29 [Removed. This was a change that I later backed out, and forgot to
608 correct the ChangeLog entry (that I had efficiently created) before
609 committing the later change.]
611 PH/30 Exim was sometimes attempting to deliver messages that had suffered
612 address errors (4xx response to RCPT) over the same connection as other
613 messages routed to the same hosts. Such deliveries are always "forced",
614 so retry times are not inspected. This resulted in far too many retries
615 for the affected addresses. The effect occurred only when there were more
616 hosts than the hosts_max_try setting in the smtp transport when it had
617 the 4xx errors. Those hosts that it had tried were not added to the list
618 of hosts for which the message was waiting, so if all were tried, there
619 was no problem. Two fixes have been applied:
621 (i) If there are any address or message errors in an SMTP delivery, none
622 of the hosts (tried or untried) are now added to the list of hosts
623 for which the message is waiting, so the message should not be a
624 candidate for sending over the same connection that was used for a
625 successful delivery of some other message. This seems entirely
626 reasonable: after all the message is NOT "waiting for some host".
627 This is so "obvious" that I'm not sure why it wasn't done
628 previously. Hope I haven't missed anything, but it can't do any
629 harm, as the worst effect is to miss an optimization.
631 (ii) If, despite (i), such a delivery is accidentally attempted, the
632 routing retry time is respected, so at least it doesn't keep
633 hammering the server.
635 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
636 in ${readsocket because some servers need this prod.
638 PH/32 Added some extra debug output when updating a wait-xxx database.
640 PH/33 The hint "could be header name not terminated by colon", which has been
641 given for certain expansion errors for a long time, was not being given
642 for the ${if def:h_colon_omitted{... case.
644 PH/34 The spec says: "With one important exception, whenever a domain list is
645 being scanned, $domain contains the subject domain." There was at least
646 one case where this was not true.
648 PH/35 The error "getsockname() failed: connection reset by peer" was being
649 written to the panic log as well as the main log, but it isn't really
650 panic-worthy as it just means the connection died rather early on. I have
651 removed the panic log writing for the ECONNRESET error when getsockname()
654 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
655 runs only) independently of the message's sender address. This meant
656 that, if the 4xx error was in fact related to the sender, a different
657 message to the same recipient with a different sender could confuse
658 things. In particualar, this can happen when sending to a greylisting
659 server, but other circumstances could also provoke similar problems.
660 I have changed the default so that the retry time for these errors is now
661 based a combination of the sender and recipient addresses. This change
662 can be overridden by setting address_retry_include_sender=false in the
665 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
666 remote server are returned as part of bounce messages. This was not
667 happening for LMTP over a pipe (the lmtp transport), but now it is the
668 same for both kinds of LMTP.
670 PH/38 Despite being documented as not happening, Exim was rewriting addresses
671 in header lines that were in fact CNAMEs. This is no longer the case.
673 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
674 and queue runs started by the daemon processed all messages. This has
675 been fixed so that -R and -S can now usefully be given with -q<time>.
677 PH/40 Import PCRE release 6.7 (fixes some bugs).
679 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
681 PH/42 Give an error if -q is specified more than once.
683 PH/43 Renamed the variables $interface_address and $interface_port as
684 $received_ip_address and $received_port, to make it clear that these
685 values apply to message reception, and not to the outgoing interface when
686 a message is delivered. (The old names remain recognized, of course.)
688 PH/44 There was no timeout on the connect() call when using a Unix domain
689 socket in the ${readsocket expansion. There now is.
691 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
692 be meaningful with "accept".
694 SC/01 Eximstats V1.43
695 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
697 SC/02 Eximstats V1.44
698 Use a glob alias rather than an array ref in the generated
699 parser. This improves both readability and performance.
701 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
702 Collect SpamAssassin and rejection statistics.
703 Don't display local sender or destination tables unless
704 there is data to show.
705 Added average volumes into the top table text output.
707 SC/04 Eximstats V1.46
708 Collect data on the number of addresses (recipients)
709 as well as the number of messages.
711 SC/05 Eximstats V1.47
712 Added 'Message too big' to the list of mail rejection
713 reasons (thanks to Marco Gaiarin).
715 SC/06 Eximstats V1.48
716 Mainlog lines which have GMT offsets and are too short to
717 have a flag are now skipped.
719 SC/07 Eximstats V1.49 (Alain Williams)
720 Added the -emptyok flag.
722 SC/08 Eximstats V1.50
723 Fixes for obtaining the IP address from reject messages.
725 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
726 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
727 whitesspace changes from 4.64-PH/27
729 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
732 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
733 are found, allow negative numbers in numeric criteria)
735 JJ/06 exipick.20061117.2, added new $message_body_missing variable
737 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
738 to match changes made in 4.64-PH/43
740 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
742 PH/47 Put in an explicit test for a DNS lookup of an address record where the
743 "domain" is actually an IP address, and force a failure. This locks out
744 those revolvers/nameservers that support "A-for-A" lookups, in
745 contravention of the specifications.
747 PH/48 When a host name was looked up from an IP address, and the subsequent
748 forward lookup of the name timed out, the host name was left in
749 $sender_host_name, contrary to the specification.
751 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
752 restricted to single-key lookups, Exim was not diagnosing an error if
753 * or *@ was used with a query-style lookup.
755 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
757 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
758 long ago, but noone interested enough thought of it. Let's just say that
759 the "1.1" means that there are some new functions that weren't there at
760 some point in the past.
762 PH/51 Error processing for expansion failure of helo_data from an smtp
763 transport during callout processing was broken.
765 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
766 tested/used via the -bh/-bhc/-bs options.
768 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
769 bug, fixed in subsequent PCRE releases).
771 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
772 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
774 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
780 SC/01 Use a glob alias rather than an array ref in eximstats generated
781 parser. This improves both readability and performance.
783 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
784 Don't display local sender or destination tables in eximstats unless
785 there is data to show.
786 Added average volumes into the eximstats top table text output.
788 SC/03 Collect data on the number of addresses (recipients) as well
789 as the number of messages in eximstats.
791 TF/01 Correct an error in the documentation for the redirect router. Exim
792 does (usually) call initgroups() when daemonizing.
794 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
795 with consistent privilege compared to when running as a daemon.
797 TF/03 Note in the spec that $authenticated_id is not set for local
798 submissions from trusted users.
800 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
801 Thanks to Dean Brooks <dean@iglou.com> for the patch.
803 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
804 by adding some example configuration directives to the default
805 configuration file. A little bit of work is required to uncomment the
806 directives and define how usernames and passwords are checked, but
807 there is now a framework to start from.
809 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
810 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
811 without this. I don't know how relevant this is to other LDAP libraries.
813 PH/02 Add the verb name to the "unknown ACL verb" error.
815 PH/03 Magnus Holmgren's patch for filter_prepend_home.
817 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
819 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
820 directory not expanded when it should be if an expanded home directory
821 was set for the address (which is overridden by the transport).
823 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
826 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
827 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
828 because it is too late at that time, and has no effect.
830 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
831 security issue with \' (bugzilla #107). I could not use the
832 PQescapeStringConn() function, because it needs a PGconn value as one of
835 PH/08 When testing addresses using -bt, indicate those final addresses that
836 are duplicates that would not cause an additional delivery. At least one
837 person was confused, thinking that -bt output corresponded to deliveries.
838 (Suppressing duplicates isn't a good idea as you lose the information
839 about possibly different redirections that led to the duplicates.)
841 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
842 systems where poll() doesn't work, in particular OS X.
844 PH/10 Added more information to debugging output for retry time not reached.
846 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
847 operations in malware.c.
849 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
852 PH/13 If write_rejectlog was set false when logging was sent to syslog with
853 syslog_duplication set false, log lines that would normally be written
854 both the the main log and to the reject log were not written to syslog at
857 PH/14 In the default configuration, change the use of "message" in ACL warn
858 statements to "add_header".
860 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
861 not followed by a command (e.g. "seen endif").
863 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
864 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
867 PH/17 Added extra conditions to the default value of delay_warning_condition
871 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
872 { match{$h_precedence:}{(?i)bulk|list|junk} } \
873 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
876 The Auto-Submitted: and various List- headers are standardised, whereas I
877 don't think Precedence: ever was.
879 PH/18 Refactored debugging code in route_finduser() to show more information,
880 in particular, the error code if getpwnam() issues one.
882 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
883 This is apparently needed in addition to the PH/07 change above to avoid
884 any possible encoding problems.
886 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
887 but not after initializing Perl.
889 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
890 output them only if debugging. By default they are written stderr,
891 apparently, which is not desirable.
893 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
896 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
899 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
901 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
902 authenticated or an ident call has been made. Suppress the default
903 values for $authenticated_id and $authenticated_sender (but permit -oMai
904 and -oMas) when testing with -bh.
906 PH/24 Re-jigged the order of the tests in the default configuration so that the
907 tests for valid domains and recipients precede the DNS black list and CSA
908 tests, on the grounds that those ones are more expensive.
910 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
911 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
912 This bug exists in every version of Exim that I still have, right back to
915 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
916 However, an attempt to turn on case-sensitivity in a regex key by
917 including (?-i) didn't work because the subject string was already
918 lowercased, and the effects were non-intuitive. It turns out that a
919 one-line patch can be used to allow (?-i) to work as expected.
925 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
926 other effects) broke the use of negated acl sub-conditions.
928 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
931 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
932 "Deny" causes Exim to reject the incoming connection with a 554 error.
933 Unfortunately, if there is a major crisis, such as a disk failure,
934 tcp-wrappers gives "deny", whereas what one would like would be some
935 kind of temporary error. A kludge has been added to help with this.
936 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
937 554 error is used if errno is still zero or contains ENOENT (which occurs
938 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
941 PH/03 Add -lutil to the default FreeBSD LIBS setting.
943 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
944 errors. Otherwise a message that provokes a temporary error (when other
945 messages do not) can cause a whole host to time out.
947 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
948 addresses were routed directly to files or pipes from a redirect router.
949 File deliveries just didn't batch; pipe deliveries might have suffered
952 PH/06 A failure to get a lock for a hints database would erroneously always say
953 "Failed to get write lock", even when it was really a read lock.
955 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
956 of 0600. This has been changed to use the value of the lockfile_mode
957 option (which defaults to 0600).
959 PH/08 Applied small patch from the Sieve maintainer.
961 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
962 folder from quota calculations, a direct delivery into this folder messed
963 up the contents of the maildirsize file. This was because the regex was
964 used only to exclude .Trash (or whatever) when the size of the mailbox
965 was calculated. There was no check that a delivery was happening into an
966 excluded directory. This bug has been fixed by ignoring all quota
967 processing for deliveries into excluded directories.
969 PH/10 Added the maildirfolder_create_regex option to appendfile.
975 PH/01 The code for finding all the local interface addresses on a FreeBSD
976 system running IPv6 was broken. This may well have applied to all BSD
977 systems, as well as to others that have similar system calls. The broken
978 code found IPv4 interfaces correctly, but gave incorrect values for the
979 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
980 that it would not match correctly against @[] and not recognize the IPv6
983 PH/02 The ipliteral router was not recognizing addresses of the form user@
984 [ipv6:....] because it didn't know about the "ipv6:" prefix.
986 PH/03 Added disable_ipv6.
988 PH/04 Changed $reply_address to use the raw form of the headers instead of the
989 decoded form, because it is most often used to construct To: headers
990 lines in autoreplies, and the decoded form may well be syntactically
991 invalid. However, $reply_address has leading white space removed, and all
992 newlines turned into spaces so that the autoreply transport does not
995 PH/05 If group was specified without a user on a router, and no group or user
996 was specified on a transport, the group from the router was ignored.
998 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
999 for visible compile-time settings that can be used to change these
1000 numbers, for those that want even more. Backwards compatibility with old
1001 spool files has been maintained. However, going back to a previous Exim
1002 release will lost any variables that are in spool files.
1004 PH/07 Two small changes when running in the test harness: increase delay when
1005 passing a TCP/IP connection to a new process, in case the original
1006 process has to generate a bounce, and remove special handling of
1007 127.0.0.2 (sic), which is no longer necessary.
1009 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1010 be the same on different OS.
1012 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1015 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1016 whether --show-vars was specified or not
1018 JJ/02 exipick: Added support for new ACL variable spool format introduced
1021 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1022 syntactically invalid From: or Reply-to: line, and a filter used this to
1023 generate an autoreply, and therefore failed to obtain an address for the
1024 autoreply, Exim could try to deliver to a non-existent relative file
1025 name, causing unrelated and misleading errors. What now happens is that
1026 it logs this as a hard delivery error, but does not attempt to create a
1029 PH/11 The exinext utility has a -C option for testing purposes, but although
1030 the given file was scanned by exinext itself; it wasn't being passed on
1031 when Exim was called.
1033 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1034 an end-of-file indication when reading a command response.
1036 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1037 compiled. In many other places in Exim, IPv6 addresses are always
1038 recognized, so I have changed this. It also means that IPv4 domain
1039 literals of the form [IPV4:n.n.n.n] are now always recognized.
1041 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1042 used if the router is not running as root, for example, when verifying at
1043 ACL time, or when using -bh. The debugging output from this situation was
1044 non-existent - all you got was a failure to exec. I have made two
1047 (a) Failures to set uid/gid, the current directory, or a process leader
1048 in a subprocess such as that created by queryprogram now generate
1049 suitable debugging ouput when -d is set.
1051 (b) The queryprogram router detects when it is not running as root,
1052 outputs suitable debugging information if -d is set, and then runs
1053 the subprocess without attempting to change uid/gid.
1055 PH/15 Minor change to Makefile for building test_host (undocumented testing
1058 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1059 additional section of a DNS packet that returns MX or SRV records.
1060 Instead, it always explicitly searches for A/AAAA records. This avoids
1061 major problems that occur when a DNS server includes only records of one
1062 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1063 fixed another bug: if SRV records were looked up and the corresponding
1064 address records were *not* found in the additional section, the port
1065 values from the SRV records were lost.
1067 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1068 using the correct key (the original address) when searching the retry
1069 rules in order to find which one to use for generating the retry hint.
1071 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1072 adding the default one. Similarly, if it contains a Reply-To: header, the
1073 errors_reply_to option, if set, is not used.
1075 PH/19 When calculating a retry time, Exim used to measure the "time since
1076 failure" by looking at the "first failed" field in the retry record. Now
1077 it does not use this if it is later than than the arrival time of the
1078 message. Instead it uses the arrival time. This makes for better
1079 behaviour in cases where some deliveries succeed, thus re-setting the
1080 "first failed" field. An example is a quota failure for a huge message
1081 when small messages continue to be delivered. Without this change, the
1082 "time since failure" will always be short, possible causing more frequent
1083 delivery attempts for the huge message than are intended.
1084 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1086 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1087 $1, $2, $3) because the numerical variables can be reset during some
1088 expansion items (e.g. "match"), thereby losing the authentication data.
1090 PH/21 Make -bV show the size of off_t variables so that the test suite can
1091 decide whether to run tests for quotas > 2G.
1093 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1094 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1095 filecount value is greater than 2G or if a quota value is greater than 2G
1096 on a system where the size of off_t is not greater than 4, a panic error
1099 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1100 never match. The debug and -bh output now contains an explicit error
1101 message indicating a malformed IPv4 address or mask.
1103 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1104 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1105 PH/23 above applies.
1107 PH/25 Do not write to syslog when running in the test harness. The only
1108 occasion when this arises is a failure to open the main or panic logs
1109 (for which there is an explicit test).
1111 PH/26 Added the /no_tell option to "control=freeze".
1113 PH/27 If a host name lookup failed very early in a connection, for example, if
1114 the IP address matched host_lookup and the reverse lookup yielded a name
1115 that did not have a forward lookup, an error message of the form "no IP
1116 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1117 could be logged. Now it outputs the IP address instead of "NULL".
1119 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1120 allows the sender and the authenticated sender to be set when
1121 submitting a message from within Exim. Since child_open_exim() is
1122 documented for local_scan(), the new function should be too.
1124 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1125 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1126 results in an empty string is now treated as unset.
1128 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1130 PH/31 Added sender_verify_fail logging option.
1132 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1133 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1134 tidied the source and removed it altogether.
1136 PH/33 When a queue run was abandoned because the load average was too high, a
1137 log line was always written; now it is written only if the queue_run log
1138 selector is set. In addition, the log line for abandonment now contains
1139 information about the queue run such as the pid. This is always present
1140 in "start" and "stop" lines but was omitted from the "abandon" line.
1142 PH/34 Omit spaces between a header name and the colon in the error message that
1143 is given when verify = headers_syntax fails (if there are lots of them,
1144 the message gets confusing).
1146 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1147 names, as there are now some PTR records that contain slashes. This check
1148 is only to protect against broken name servers that fall over on strange
1149 characters, so the fact that it applies to all lookups doesn't matter.
1151 PH/36 Now that the new test suite is complete, we can remove some of the
1152 special code in Exim that was needed for the old test suite. For example,
1153 sorting DNS records because real resolvers return them in an arbitrary
1154 order. The new test suite's fake resolver always returns records in the
1157 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1158 bounces) except when queue_only is set, to avoid logging races between
1159 the different processes.
1161 PH/38 Panic-die if .include specifies a non-absolute path.
1163 PH/39 A tweak to the "H" retry rule from its user.
1165 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1166 a label. They prevented compilation on older perls.
1168 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1169 a warning to be raised on newish perls.
1171 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1172 on queue. Changes to match documented behaviour of showing count of
1173 messages matching specified criteria.
1175 PH/40 Changed the default ident timeout from 30s to 5s.
1177 PH/41 Added support for the use of login_cap features, on those BSD systems
1178 that have them, for controlling the resources used by pipe deliveries.
1180 PH/42 The content-scanning code uses fopen() to create files in which to put
1181 message data. Previously it was not paying any attention to the mode of
1182 the files. Exim runs with umask(0) because the rest of the code creates
1183 files with open(), and sets the required mode explicitly. Thus, these
1184 files were ending up world-writeable. This was not a big issue, because,
1185 being within the spool directory, they were not world-accessible. I have
1186 created a function called modefopen, which takes an additional mode
1187 argument. It sets umask(777), creates the file, chmods it to the required
1188 mode, then resets the umask. All the relevant calls to fopen() in the
1189 content scanning code have been changed to use this function.
1191 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1192 to 24 hours. This avoids potential overflow problems when processing G
1193 and H retry rules. I suspect nobody ever tinkers with this value.
1195 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1197 PH/45 When the plaintext authenticator is running as a client, the server's
1198 challenges are checked to ensure they are valid base64 strings. By
1199 default, the authentication attempt is cancelled if an invalid string is
1200 received. Setting client_ignore_invalid_base64 true ignores these errors.
1201 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1202 they are received. Thus, the responses can be made to depend on the
1203 challenges. If an invalid string is ignored, an empty string is placed in
1206 PH/46 Messages that are created by the autoreply transport now contains a
1207 References: header, in accordance with RFCs 2822 and 3834.
1209 PH/47 Added authenticated_sender_force to the smtp transport.
1211 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1213 PH/49 Installed latest patch from the Sieve maintainer.
1215 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1216 was also set, the appendfile transport was unnecessarily scanning a
1217 directory of message files (e.g. for maildir delivery) to find the count
1218 of files (along with the size), even though it did not need this
1219 information. It now does the scan only if it needs to find either the
1220 size of the count of files.
1222 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1224 PH/52 Two bugs concerned with error handling when the smtp transport is
1227 (i) Exim was not creating retry information for temporary errors given
1228 for individual recipients after the DATA command when the smtp transport
1229 was used in LMTP mode. This meant that they could be retried too
1230 frequently, and not timed out correctly.
1232 (ii) Exim was setting the flag that allows error details to be returned
1233 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1234 recipients that were returned after the DATA command.
1236 PH/53 This is related to PH/52, but is more general: for any failing address,
1237 when detailed error information was permitted to be returned to the
1238 sender, but the error was temporary, then after the final timeout, only
1239 "retry timeout exceeded" was returned. Now it returns the full error as
1240 well as "retry timeout exceeded".
1242 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1243 do this, and (what is worse) MTAs that accept it.
1245 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1246 will now be deprecated.
1248 PH/56 New os.c-cygwin from the Cygwin maintainer.
1250 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1251 formats (previously only available in exim formats via -bpr, -bpru,
1252 and -bpra. Now also available in native and exiqgrep formats)
1254 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1255 with very large, slow to parse queues
1257 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1259 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1261 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1262 responses to authentication challenges, though it was showing the
1263 challenges; (ii) I've removed the CR characters from the debug output for
1266 PH/58 Allow for the insertion of a newline as well as a space when a string
1267 is turned into more than one encoded-word during RFC 2047 encoding. The
1268 Sieve code now uses this.
1270 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1271 data_4xx, lost_connection, tls_required.
1273 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1274 message was being sent as an SMTP response.
1276 PH/61 Add -l and -k options to exicyclog.
1278 PH/62 When verifying, if an address was redirected to one new address, so that
1279 verification continued, and the new address failed or deferred after
1280 having set something in $address_data, the value of $address_data was not
1281 passed back to the ACL. This was different to the case when no
1282 redirection occurred. The value is now passed back in both cases.
1284 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1285 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1286 use login_cap.h, so on its own it isn't the distinguishing feature. The
1287 new name refers directly to the setclassresources() function.
1289 PH/65 Added configuration files for NetBSD3.
1291 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1293 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1294 is preferred over IPv4.
1296 PH/68 The bounce_return_message and bounce_return_body options were not being
1297 honoured for bounces generated during the reception of non-SMTP messages.
1298 In particular, this applied to messages rejected by the ACL. This bug has
1299 been fixed. However, if bounce_return_message is true and bounce_return_
1300 body is false, the headers that are returned for a non-SMTP message
1301 include only those that have been read before the error was detected.
1302 (In the case of an ACL rejection, they have all been read.)
1304 PH/69 The HTML version of the specification is now built in a directory called
1305 spec_html instead of spec.html, because the latter looks like a path with
1306 a MIME-type, and this confuses some software.
1308 PH/70 Catch two compiler warnings in sieve.c.
1310 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1311 function verify_get_ident() calls ip_connect() to connect a socket, but
1312 if the "connect()" function timed out, ip_connect() used to close the
1313 socket. However, verify_get_ident() also closes the socket later, and in
1314 between Exim writes to the log, which may get opened at this point. When
1315 the socket was closed in ip_connect(), the log could get the same file
1316 descriptor number as the socket. This naturally causes chaos. The fix is
1317 not to close the socket in ip_connect(); the socket should be closed by
1318 the function that creates it. There was only one place in the code where
1319 this was missing, in the iplookup router, which I don't think anybody now
1320 uses, but I've fixed it anyway.
1322 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1323 well as to direct DNS lookups. Otherwise the handling of names in host
1324 lists is inconsistent and therefore confusing.
1330 PH/01 Two changes to the default runtime configuration:
1332 (1) Move the checks for relay_from_hosts and authenticated clients from
1333 after to before the (commented out) DNS black list checks.
1335 (2) Add control=submission to the relay_from_hosts and authenticated
1336 clients checks, on the grounds that messages accepted by these
1337 statements are most likely to be submissions.
1339 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1341 (1) Generate an error if the third argument for the ${prvs expansion is
1344 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1347 (3) Reset the variables that are obtained from the first argument of
1348 ${prvscheck and used in the second argument before leaving the code,
1349 because their memory is reclaimed, so using them afterwards may do
1352 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1353 one (it's much easier than Tom thought :-).
1355 (5) Because of (4), we can now allow for the use of $prvscheck_result
1356 inside the third argument.
1358 PH/03 For some reason, the default setting of PATH when running a command from
1359 a pipe transport was just "/usr/bin". I have changed it to
1362 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1363 anything to be listed in the output from -bV.
1365 PH/05 When a filter generated an autoreply, the entire To: header line was
1366 quoted in the delivery log line, like this:
1368 => >A.N.Other <ano@some.domain> <original@ddress> ...
1370 This has been changed so that it extracts the operative address. There
1371 may be more than one such address. If so, they are comma-separated, like
1374 => >ano@some.domain,ona@other.domain <original@ddress> ...
1376 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1377 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1378 not being looked up in the rDNS to get a host name, Exim was showing the
1379 IP address twice in Received: lines, even though the IP addresses were
1380 identical. For example:
1382 Received: from [1.2.3.4] (helo=[1.2.3.4])
1384 However, if the real host name was known, it was omitting the HELO data
1385 if it matched the actual IP address. This has been tidied up so that it
1386 doesn't show the same IP address twice.
1388 PH/07 When both +timestamp and +memory debugging was on, the value given by
1389 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1390 called by the debug printing, thereby overwriting the timestamp buffer.
1391 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1393 PH/08 When the original message was included in an autoreply transport, it
1394 always said "this is a copy of the message, including all the headers",
1395 even if body_only or headers_only was set. It now gives an appropriate
1398 PH/09 Applied a patch from the Sieve maintainer which:
1400 o fixes some comments
1401 o adds the (disabled) notify extension core
1402 o adds some debug output for the result of if/elsif tests
1403 o points to the current vacation draft in the documentation
1404 and documents the missing references header update
1408 o fixes a bug in processing the envelope test (when testing
1409 multiple envelope elements, the last element determinted the
1412 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1413 Electronic Mail") by including:
1415 Auto-submitted: auto-generated
1417 in the messages that it generates (bounce messages and others, such as
1418 warnings). In the case of bounce messages for non-SMTP mesages, there was
1419 also a typo: it was using "Auto_submitted" (underscore instead of
1420 hyphen). Since every message generated by Exim is necessarily in response
1421 to another message, thes have all been changed to:
1423 Auto-Submitted: auto-replied
1425 in accordance with these statements in the RFC:
1427 The auto-replied keyword:
1429 - SHOULD be used on messages sent in direct response to another
1430 message by an automatic process,
1432 - MUST NOT be used on manually-generated messages,
1434 - MAY be used on Delivery Status Notifications (DSNs) and Message
1435 Disposition Notifications (MDNs),
1437 - MUST NOT be used on messages generated by automatic or periodic
1438 processes, except for messages which are automatic responses to
1441 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1442 to the default Received: header definition.
1444 PH/12 Added log selector acl_warn_skipped (default on).
1446 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1447 variables because (a) they are in the wrong storage pool and (b) even if
1448 they were copied, it wouldn't work properly because of the caching.
1450 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1451 checking when decoding. Apparently there are clients that generate
1452 overlong encoded strings. Why am I not surprised?
1454 PH/15 If the first argument of "${if match_address" was not empty, but did not
1455 contain an "@" character, Exim crashed. Now it writes a panic log message
1456 and treats the condition as false.
1458 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1460 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1461 extension "envelope-auth". The code is finished and in agreement with
1462 other implementations, but there is no documentation so far and in fact,
1463 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1464 not changing the active code.
1466 Print executed "if" and "elsif" statements when debugging is used. This
1467 helps a great deal to understand what a filter does.
1469 Document more things not specified clearly in RFC3028. I had all this
1470 sorted out, when out of a sudden new issues came to my mind. Oops."
1472 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1475 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1476 canonical form (as documented). However, after a host name lookup from
1477 the IP address, check_host() was doing a simple string comparison with
1478 addresses acquired from the DNS when checking that the found name did
1479 have the original IP as one of its addresses. Since any found IPv6
1480 addresses are likely to be in abbreviated form, the comparison could
1481 fail. Luckily, there already exists a function for doing the comparison
1482 by converting both addresses to binary, so now that is used instead of
1483 the text comparison.
1485 PH/20 There was another similar case to PH/19, when a complete host name was
1486 given in a host list; looking up its IP address could give an abbreviated
1487 form, whereas the current host's name might or might not be abbreviated.
1488 The same fix has been applied.
1494 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1495 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1498 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1499 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1501 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1503 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1504 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1505 Also, for "Auto-Submitted", treat anything other than "no" as
1506 non-personal, in accordance with RFC 3834. (Previously it treated
1507 anything starting "auto-" as non-personal.)
1509 TF/01 The control=submission/name=... option had a problem with syntax
1510 errors if the name included a slash character. The /name= option
1511 now slurps the rest of the string, so it can include any characters
1512 but it must come last in the list of options (after /sender_retain
1515 PH/05 Some modifications to the interface to the fake nameserver for the new
1523 TK/01 Added the "success_on_redirect" address verification option. See
1524 NewStuff for rationale and an example.
1526 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1528 PH/02 Patch to exigrep to allow it to work on syslog lines.
1530 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1531 fread() to skip over the body file's header line, because in Cygwin the
1532 header line is locked and is inaccessible.
1534 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1535 co-exist for some time) to make it clear that it is the Exim ID that is
1536 referenced, not the Message-ID: header line.
1538 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1539 string_format() function, because snprintf() does not exist on all
1542 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1543 ${stat: expansion item.
1545 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1546 protocol synchronization error", to keep the pedants happy.
1548 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1549 well as for IRIX systems, when gcc is being used. See the host.c source
1552 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1554 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1557 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1558 transport and to the smtp transport in LMTP mode.
1560 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1562 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1563 run a filter in a subprocess. This could lead to confusion in subsequent
1564 lookups in the parent process. There should also be a search_tidyup() at
1565 the end of the subprocess.
1567 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1568 only if the host matched helo_try_verify_hosts, which caused the
1569 verification to occur when the EHLO/HELO command was issued. The ACL just
1570 tested the remembered result. Now, if a previous verification attempt has
1571 not happened, "verify = helo" does it there and then.
1573 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1575 TK/03 Fix log output including CR from clamd.
1577 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1578 exist provoked a memory error which could cause a segfault.
1580 PH/15 Installed PCRE 6.2
1582 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1584 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1585 of the problem. Specifically, suggested +O2 rather than +O1 for the
1588 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1590 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1591 record, the port was not correctly logged when the outgoing_port log
1592 selector was set (it logged the transort's default port).
1594 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1595 fallback_hosts, and "hosts" in the smtp transport.
1597 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1598 host errors such as "Connection refused".
1600 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1601 authentication with radiusclient 0.4.9:
1603 - Error returned from rc_read_config was caught wrongly
1604 - Username/password not passed on to radius server due to wrong length.
1606 The presumption is that some radiusclient API changes for 4.51/PH/17
1607 were not taken care of correctly. The code is still untested by me (my
1608 Linux distribution still has 0.3.2 of radiusclient), but it was
1609 contributed by a Radius user.
1611 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1612 expanding the "port" option of the smtp transport.
1614 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1615 while reading a MIME header. Thanks to Tom Hughes for a patch.
1617 PH/24 Include config.h inside local_scan.h so that configuration settings are
1620 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1621 that in an ACL for RCPT (for example), you can examine exactly what was
1624 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1625 commands, but it was not correctly comparing the address with the actual
1626 client host address. Thus, it would show the EHLO address in Received:
1627 header lines when this was not necessary.
1629 PH/27 Added the % operator to ${eval:}.
1631 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1632 it should be ignoring failures (because with -C, for example, it has lost
1633 privilege). It wasn't ignoring creation failures other than "already
1636 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1639 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1640 because an input line was too long, either on its own, or by virtue of
1641 too many continuations, the temporary file was not being removed, and the
1642 return code was incorrect.
1644 PH/31 Missing "BOOL" in function definition in filtertest.c.
1646 PH/32 Applied Sieve patches from the maintainer.
1648 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1650 PH/33 Added "verify = not_blind".
1652 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1653 Local/Makefile (with some defaults set). These are used in built scripts
1654 such as exicyclog, but they have never been used in the exim_install
1655 script (though there are many overriding facilities there). I have
1656 arranged that the exim_install script now takes note of these two
1659 PH/35 Installed configuration files for Dragonfly.
1661 PH/36 When a locally submitted message by a trusted user did not contain a
1662 From: header, and the sender address was obtained from -f or from an SMTP
1663 MAIL command, and the trusted user did not use -F to supply a sender
1664 name, $originator_name was incorrectly used when constructing a From:
1665 header. Furthermore, $originator_name was used for submission mode
1666 messages from external hosts without From: headers in a similar way,
1667 which is clearly wrong.
1669 PH/37 Added control=suppress_local_fixups.
1671 PH/38 When log_selector = +received_sender was set, and the addition of the
1672 sender made the log line's construction buffer exactly full, or one byte
1673 less than full, an overflow happened when the terminating "\n" was
1676 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1677 when the result of a list match is failure because a DNS lookup failed.
1679 PH/40 RM_COMMAND is now used in the building process.
1681 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1682 the "build-* directories that it finds.
1684 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1685 address in a domain literal was a prefix of an interface address.
1687 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1688 when verifying a sender address, unless rewrite_headers is false.
1690 PH/44 Wrote a long comment about why errors_to addresses are verified as
1691 recipients, not senders.
1693 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1694 the ratelimit ACL was added.
1696 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1698 PH/46 Added extra information about PostgreSQL errors to the error string.
1700 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1701 avoiding the need to install special zones in a real server. This is
1702 backwards compatible; if it can't find the fake resolver, it drops back.
1703 Thus, both old and new test suites can be run.
1705 TF/02 Added util/ratelimit.pl
1707 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1708 clock is set back in time.
1710 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1711 Candler <B.Candler@pobox.com>.
1713 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1714 OK for addresses that are the result of redirections.
1716 PH/48 A number of further additions for the benefit of the new test suite,
1717 including a fake gethostbyname() that interfaces to the fake DNS resolver
1720 TF/06 The fix for widen_domains has also been applied to qualify_single and
1721 search_parents which are the other dnslookup options that can cause
1724 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1727 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1729 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1730 service" if it encountered an MX record with an empty target hostname.
1731 The message is now "an MX or SRV record indicated no SMTP service".
1733 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1734 if the DNS of the sending site is misconfigured. This is quite a
1735 common situation. This change restores the behaviour of treating a
1736 helo verification defer as a failure.
1738 PH/51 If self=fail was set on a router, the bounce message did not include the
1739 actual error message.
1745 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1747 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1748 command itself ended in error, the underlying message about the transport
1749 filter timeout was being overwritten with the pipe command error. Now the
1750 underlying error message should be appended to the second error message.
1752 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1754 PH/02 Reduce the amount of output that "make" produces by default. Full output
1755 can still be requested.
1757 PH/03 The warning log line about a condition test deferring for a "warn" verb
1758 was being output only once per connection, rather than after each
1759 occurrence (because it was using the same function as for successful
1760 "warn" verbs). This seems wrong, so I have changed it.
1762 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1763 it should not have, which might have caused a crash in the right
1764 circumstances, but probably never did.
1766 PH/04 Installed a modified version of Tony Finch's patch to make submission
1767 mode fix the return path as well as the Sender: header line, and to
1768 add a /name= option so that you can make the user's friendly name appear
1771 TF/03 Added the control = fakedefer ACL modifier.
1773 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1774 Mark Lowes for thorough testing.
1776 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1778 TK/03 Merged latest SRS patch from Miles Wilton.
1780 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1781 with the definition in sysexits.h (which is #included earlier).
1782 Fortunately, Exim does not actually use EX_OK. The code used to try to
1783 preserve the sysexits.h value, by assumimg that macro definitions were
1784 scanned for macro replacements. I have been disabused of this notion,
1785 so now the code just undefines EX_OK before #including unistd.h.
1787 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1788 in the smtp transport. When a block could not be written in a single
1789 write() function, the timeout was being re-applied to each part-write.
1790 This seems wrong - if the receiver was accepting one byte at a time it
1791 would take for ever. The timeout is now adjusted when this happens. It
1792 doesn't have to be particularly precise.
1794 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1795 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1797 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1798 without a domain if the check to <postmaster@domain> fails.
1800 SC/01 Eximstats: added -xls and the ability to specify output files
1801 (patch written by Frank Heydlauf).
1803 SC/02 Eximstats: use FileHandles for outputing results.
1805 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1807 SC/04 Eximstats: fixed display of large numbers with -nvr option
1809 SC/05 Eximstats: fixed merging of reports with empty tables.
1811 SC/06 Eximstats: added the -include_original_destination flag
1813 SC/07 Eximstats: removed tabs and trailing whitespace.
1815 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1817 TK/06 MBOX spool code: Add real "From " MBOX separator line
1818 so the .eml file is really in mbox format (even though
1819 most programs do not really care). Patch from Alex Miller.
1821 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1822 The latter is generated from $received_to and is only set if the
1823 message has one envelope recipient. SA can use these headers,
1824 obviously out-of-the-box. Patch from Alex Miller.
1826 PH/08 The ${def test on a variable was returning false if the variable's
1827 value was "0", contrary to what the specification has always said!
1828 The result should be true unless the variable is empty.
1830 PH/09 The syntax error of a character other than { following "${if
1831 def:variable_name" (after optional whitespace) was not being diagnosed.
1832 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1833 accidental colon was present, for example, could give incorrect results.
1835 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1836 result is used (not including appendfile, where other changes are about
1839 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1840 This involved changing a lot of size variables from int to off_t. It
1841 should work with maildirs and everything.
1843 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1844 spamd dying while we are connected to it.
1846 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1849 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1850 function for BDB 4.3.
1852 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1854 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1855 the sources for PCRE have been re-arranged and more files are now
1858 PH/15 The code I had for printing potentially long long variables in PH/11
1859 above was not the best (it lost precision). The length of off_t variables
1860 is now inspected at build time, and an appropriate printing format (%ld
1861 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1862 to be "long long int" or "long int". This is needed for the internal
1863 formatting function string_vformat().
1865 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1866 the configuration file to be ":syslog", then the script "guesses" where
1867 the logs files are, rather than using the compiled in default. In our
1868 case the guess is not the same as the compiled default, so the script
1869 suddenly stopped working when I started to use syslog. The patch checks
1870 to see if log_file_path is "". If so, it attempts to read it from exim
1871 with no configuration file to get the compiled in version, before it
1872 falls back to the previous guessing code."
1874 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1875 implementing BATV in an Exim configuration. See NewStuff for the gory
1878 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1879 Makefile that are specific to HP-UX.
1881 PH/18 If the "use_postmaster" option was set for a recipient callout together
1882 with the "random" option, the postmaster address was used as the MAIL
1883 FROM address for the random test, but not for the subsequent recipient
1884 test. It is now used for both.
1886 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1887 patch removes a few documentation additions to RFC 3028, because the
1888 latest draft now contains them. It adds the new en;ascii-case comparator
1889 and a new error check for 8bit text in MIME parts. Comparator and
1890 require names are now matched exactly. I enabled the subaddress
1891 extension, but it is not well tested yet (read: it works for me)."
1893 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1894 rework some of the code of TK/09 above to avoid the hardwired use of
1895 "%lld" and "long long". Replaced the call to snprintf() with a call to
1898 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1899 records point to non-existent hosts", "retry timeout exceeded", and
1900 "retry time not reached for any host after a long failure period".
1902 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1903 experimental DomainKeys support:
1905 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1906 (2) On an error such as an illegally used "control", the wrong name for
1907 the control was given.
1909 These problems did NOT occur unless DomainKeys support was compiled.
1911 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1913 PH/24 Added ${if match_ip condition.
1915 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1916 databases so that it will be absolutely obvious if a crash occurs in the
1917 DB library. This is a regular occurrence (often caused by mis-matched
1920 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1921 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1922 picked up on a user's system that detects such things. There doesn't seem
1923 to be a gcc warning option for this - only an attribute that has to be
1924 put on the function's prototype. It seems that in Fedora Core 4 they have
1925 set this on a number of new functions. No doubt there will be more in due
1928 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1929 specify a transport. However, if an address that was verified by such a
1930 router was the subject of a callout, Exim crashed because it tried to
1931 read the rcpt_include_affixes from the non-existent transport. Now it
1932 just assumes that the setting of that option is false. This bug was
1933 introduced by 4.51/PH/31.
1935 PH/28 Changed -d+all to exclude +memory, because that information is very
1936 rarely of interest, but it makes the output a lot bigger. People tend to
1937 do -d+all out of habit.
1939 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1940 code in os-type was giving problems when libc.so lives in lib64, like on
1943 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1944 aren't the modern standard, and it seems that some systems' include files
1945 don't always have them. Exim was already checking for some of the newer
1946 ones like T_AAAA, and defining it itself. I've added checks for all the
1947 record types that Exim uses.
1949 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1950 not automatically generating a new one, as it is supposed to. This
1951 prevented TLS from working. If the file did exist, but contained invalid
1952 data, a new version was generated, as expected. It was only the case of a
1953 non-existent file that was broken.
1955 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1956 with a change in libdomainkeys > 0.64.
1958 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1959 from DNS. If the selector record carries the flag, it now has
1960 precedence over the domain-wide flag.
1962 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1964 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1965 the use of a port name that isn't defined in /etc/services) occurred, the
1966 message was deferred as in a normal delivery, and thus remained on the
1967 spool, instead of being failed because of the mua_wrapper setting. This
1968 is now fixed, and I tidied up some of the mua_wrapper messages at the
1971 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1972 the messages in a hash of arrays rather than using individual hashes.
1973 This is a bit cleaner and results in dramatic memory savings, albeit
1974 at a slight CPU cost.
1976 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1977 as requested by Marc Sherman.
1979 SC/10 Eximstats: added histograms for user specified patterns as requested
1982 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1984 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1985 fopen() in the content-scanning modules that did not already have it.
1991 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1992 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1994 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1996 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1998 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1999 to test Sieve filters that use "vacation".
2001 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2002 that changes the way the GnuTLS parameters are stored in the cache file.
2003 The new format can be generated externally. For backward compatibility,
2004 if the data in the cache doesn't make sense, Exim assumes it has read an
2005 old-format file, and it generates new data and writes a new file. This
2006 means that you can't go back to an older release without removing the
2009 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2010 work if there are any delivery delays because "one_time" forces the
2011 parent to be marked "delivered", so its unseen clone is never tried
2012 again. For this reason, Exim now forbids the simultaneous setting of
2015 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2016 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2017 entry if you want to know the details. The fix, however, neglected to
2018 consider the case when local delivery batching is involved. The test for
2019 "previously delivered" was not happening when checking to see if an
2020 address could be batched with a previous (undelivered) one; under
2021 certain circumstances this could lead to multiple deliveries to the same
2024 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2025 in its include files, and this causes problems building Exim.
2027 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2028 header_syntax) but Exim was just ignoring anything given after a slash.
2029 In particular, this caused confusion with an attempt to use "verify =
2030 reverse_host_lookup/defer_ok". An error is now given when options are
2031 supplied for verify items that do not have them. (Maybe reverse_host_
2032 lookup should have a defer_ok option, but that's a different point.)
2034 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2035 defined by RFC 821) to 2048, because there were problems with some AUTH
2036 commands, and RFC 1869 says the size should be increased for extended
2037 SMTP commands that take arguments.
2039 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2042 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2043 "unknown" error; now it says that the functionality isn't in the binary.
2045 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2046 an address' error message when a string expansion fails (syntax or
2047 whatever). Otherwise the password may appear in the log. Following change
2048 PH/42 below, there is no longer a chance of it appearing in a bounce
2051 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2053 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2054 of its addresses was ever tried. (Bugzilla bug #2.)
2056 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2057 the result incorrectly in the debug output. (It correctly added a newline
2058 to what was transported.)
2060 TF/01 Added $received_time.
2062 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2063 commented out examples of how to interface to a virus scanner and to
2064 SpamAssassin. Also added commented examples of av_scanner and
2065 spamd_address settings.
2067 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2068 and controls are allowed in which ACLs. There were a couple of minor
2069 errors. Some of the entries in the conditions table (which is a table of
2070 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2071 negation of where the condition IS allowed.
2073 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2075 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2076 header file does not have a version number, so I've had to invent a new
2077 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2078 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2079 radiusclient), but it was contributed by a Radius user.
2081 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2082 files or directories, for OpenSSL.
2084 PH/19 When an Exim process that is running as root has to create an Exim log
2085 file, it does so in a subprocess that runs as exim:exim so as to get the
2086 ownership right at creation (otherwise, other Exim processes might see
2087 the file with the wrong ownership). There was no test for failure of this
2088 fork() call, which would lead to the process getting stuck as it waited
2089 for a non-existent subprocess. Forks do occasionally fail when resources
2090 run out. I reviewed all the other calls to fork(); they all seem to check
2093 PH/20 When checking for unexpected SMTP input at connect time (before writing
2094 the banner), Exim was not dealing correctly with a non-positive return
2095 from the read() function. If the client had disconnected by this time,
2096 the result was a log entry for a synchronization error with an empty
2097 string after "input=" when read() returned zero. If read() returned -1
2098 (an event I could not check), uninitialized data bytes were printed.
2099 There were reports of junk text (parts of files, etc) appearing after
2102 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2104 PH/22 Added support for macro redefinition, and (re)definition in between
2105 driver and ACL definitions.
2107 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2108 forgetting to use the resulting value; it was using the unexpanded value.
2110 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2111 hadn't been configured. The fix is from Juergen Kreileder, who
2112 understands it better than I do:
2114 "Here's what I see happening with three configured cyrus_sasl
2115 authenticators configured (plain, login, cram-md5):
2117 On startup auth_cyrus_sasl_init() gets called for each of these.
2118 This means three calls to sasl_listmech() without a specified mech_list.
2119 => SASL tests which mechs of all available mechs actually work
2120 => three warnings about OTP not working
2121 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2123 With the patch, sasl_listmech() also gets called three times. But now
2124 SASL's mech_list option is set to the server_mech specified in the the
2125 authenticator. Or in other words, the answer from sasl_listmech()
2126 gets limited to just the mech you're testing for (which is different
2128 => the return list contains just 'plain' or 'login', 'cram-md5' or
2129 nothing depending on the value of ob->server_mech.
2131 I've just tested the patch: Authentication still works fine,
2132 unavailable mechs specified in the exim configuration are still
2133 caught, and the auth.log warnings about OTP are gone."
2135 PH/25 When debugging is enabled, the contents of the command line are added
2136 to the debugging output, even when log_selector=+arguments is not
2139 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2140 answer is "GNU", and only if the return is "GNU/something" is the answer
2143 PH/27 $acl_verify_message is now set immediately after the failure of a
2144 verification in an ACL, and so is available in subsequent modifiers. In
2145 particular, the message can be preserved by coding like this:
2147 warn !verify = sender
2148 set acl_m0 = $acl_verify_message
2150 Previously, $acl_verify_message was set only while expanding "message"
2151 and "log_message" when a very denied access.
2153 PH/28 Modified OS/os.c-Linux with
2155 -#ifndef OS_LOAD_AVERAGE
2156 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2158 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2159 nomenclature these days.)
2161 PH/29 Installed patch from the Sieve maintainer that adds the options
2162 sieve_useraddress and sieve_subaddress to the redirect router.
2164 PH/30 In these circumstances:
2165 . Two addresses routed to the same list of hosts;
2166 . First host does not offer TLS;
2167 . First host accepts first address;
2168 . First host gives temporary error to second address;
2169 . Second host offers TLS and a TLS session is established;
2170 . Second host accepts second address.
2171 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2172 and peerdn, if requested) that were in fact used only for the second
2175 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2176 attention to any local part prefix or suffix that was matched by the
2177 router that accepted the address. It now behaves in the same way as it
2178 does for delivery: the affixes are removed from the local part unless
2179 rcpt_include_affixes is set on the transport.
2181 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2182 timeout during the DATA phase of an incoming message.
2184 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2185 applied a patch sanctioned by the Sieve maintainer.
2187 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2188 the uid or gid is negative. A case of a negative gid caused this to be
2189 noticed. The fix allows for either to be negative.
2191 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2192 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2195 PH/36 The change PH/12 above was broken. Fixed it.
2197 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2198 the grounds that routing the same address twice would always produce the
2199 same answer. This might have been true once, but it is certainly no
2200 longer true now. Routing a child address may depend on the previous
2201 routing that produced that child. Some complicated redirection strategies
2202 went wrong when messages had multiple recipients, and made Exim's
2203 behaviour dependent on the order in which the addresses were given.
2205 I have moved the duplicate checking until after the routing is complete.
2206 Exim scans the addresses that are assigned to local and remote
2207 transports, and removes any duplicates. This means that more work will be
2208 done, as duplicates will always all be routed, but duplicates are
2209 presumably rare, so I don't expect this is of any significance.
2211 For deliveries to pipes, files, and autoreplies, the duplicate checking
2212 still happens during the routing process, since they are not going to be
2215 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2216 It corrects a timeout issue with spamd. This is Ian's comment: "The
2217 background is that sometimes spamd either never reads data from a
2218 connection it has accepted, or it never writes response data. The exiscan
2219 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2220 blindly assumes that writes won't block so it may never time out."
2222 PH/39 Allow G after quota size as well as K and M.
2224 PH/40 The value set for $authenticated_id in an authenticator may not contain
2225 binary zeroes or newlines because the value is written to log lines and
2226 to spool files. There was no check on this. Now the value is run through
2227 the string_printing() function so that such characters are converted to
2228 printable escape sequences.
2230 PH/41 $message_linecount is a new variable that contains the total number of
2231 lines in the message. Compare $body_linecount, which is the count for the
2234 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2235 bounce and delay warning messages, except in certain special cases, which
2238 (a) An SMTP error message from a remote host;
2239 (b) A message specified in a :fail: redirection;
2240 (c) A message specified in a "fail" command in a system filter;
2241 (d) A message specified in a FAIL return from the queryprogram router;
2242 (e) A message specified by the cannot_route_message router option.
2244 In these cases only, Exim does include the error details in bounce and
2245 warning messages. There are also a few cases where bland messages such
2246 as "unrouteable address" or "local delivery error" are given.
2248 PH/43 $value is now also set for the "else" part of a ${run expansion.
2250 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2251 being worked on, but at least Exim now implements the latest version to
2254 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2255 process to complete was treated as a delivery failure, a timeout while
2256 writing the message to the pipe was logged, but erroneously treated as a
2257 successful delivery. Such timeouts include transport filter timeouts. For
2258 consistency with the overall process timeout, these timeouts are now
2259 treated as errors, giving rise to delivery failures by default. However,
2260 there is now a new Boolean option for the pipe transport called
2261 timeout_defer, which, if set TRUE, converts the failures into defers for
2262 both kinds of timeout. A transport filter timeout is now identified in
2265 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2266 systems where "make" and "gmake" are different, calling "gmake" at top
2267 level broke things. I've arranged for the value of $(MAKE) to be passed
2268 from the Makefile to this script so that it can call the same version of
2272 A note about Exim versions 4.44 and 4.50
2273 ----------------------------------------
2275 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2276 changes of various kinds. As a consequence, a big documentation update was
2277 needed. This delayed the release for rather longer than seemed good, especially
2278 in the light of a couple of (minor) security issues. Therefore, the changes
2279 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2280 release. So 4.44 and 4.50 are in effect two different branches that both start
2283 I have left the 4.50 change log unchanged; it contains all the changes since
2284 4.43. The change log for 4.44 is below; many of its items are identical to
2285 those for 4.50. This seems to be the most sensible way to preserve the
2286 historical information.
2292 1. Minor wording change to the doc/README.SIEVE file.
2294 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2295 computation of the current number of files was incorrect.
2297 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2298 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2301 4. Give more explanation in the error message when the command for a transport
2302 filter fails to execute.
2304 5. There are several places where Exim runs a non-Exim command in a
2305 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2306 was being done only for the command run by the queryprogram router. It is
2307 now done for all such subprocesses. The other cases are: ${run, transport
2308 filters, and the commands run by the lmtp and pipe transports.
2310 6. Added CONFIGURE_GROUP build-time option.
2312 7. Some older OS have a limit of 256 on the maximum number of file
2313 descriptors. Exim was using setrlimit() to set 1000 as a large value
2314 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2315 systems. I've change it so that if it can't get 1000, it tries for 256.
2317 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2318 was an oversight, and furthermore, ever since the addition of extra
2319 controls (e.g. 4.43/32), the checks on when to allow different forms of
2320 "control" were broken. There should now be diagnostics for all cases when a
2321 control that does not make sense is encountered.
2323 9. Added the /retain_sender option to "control=submission".
2325 10. $recipients is now available in the predata ACL (oversight).
2327 11. Tidy the search cache before the fork to do a delivery from a message
2328 received from the command line. Otherwise the child will trigger a lookup
2329 failure and thereby defer the delivery if it tries to use (for example) a
2330 cached ldap connection that the parent has called unbind on.
2332 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2333 of $address_data from the recipient verification was clobbered by the
2334 sender verification.
2336 13. The value of address_data from a sender verification is now available in
2337 $sender_address_data in subsequent conditions in the ACL statement.
2339 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2341 15. Added a new option "connect=<time>" to callout options, to set a different
2344 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2345 was its contents. (It was OK if the option was not defined at all.)
2347 17. A "Completed" log line is now written for messages that are removed from
2348 the spool by the -Mrm option.
2350 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2351 information about exactly what failed.
2353 19. Added -dd to debug only the daemon process.
2355 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2356 handles timeouts, both on the server side and network timeouts. Renamed the
2357 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2359 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2360 It is now set to "smtps".
2362 22. $host_address is now set to the target address during the checking of
2363 ignore_target_hosts.
2365 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2366 being passed; this would have caused $sender_host_name to have been used if
2367 matching the list had actually called for a host name (not very likely,
2368 since this list is usually IP addresses). A host name is now passed as
2371 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2372 code that specifies a non-restarting handler (typically sigaction() in
2373 modern systems) in an attempt to fix a rare and obscure crash bug.
2375 25. Narrowed the window for a race in the daemon that could cause it to ignore
2376 SIGCHLD signals. This is not a major problem, because they are used only to
2377 wake it up if nothing else does.
2379 26. A malformed maildirsize file could cause Exim to calculate negative values
2380 for the mailbox size or file count. Odd effects could occur as a result.
2381 The maildirsize information is now recalculated if the size or filecount
2384 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2385 support for a long time. Removed HAVE_SYS_VFS_H.
2387 28. Installed the latest version of exipick from John Jetmore.
2389 29. In an address list, if the pattern was not a regular expression, an empty
2390 subject address (from a bounce message) matched only if the pattern was an
2391 empty string. Non-empty patterns were not even tested. This was the wrong
2392 because it is perfectly reasonable to use an empty address as part of a
2393 database query. An empty address is now tested by patterns that are
2394 lookups. However, all the other forms of pattern expect the subject to
2395 contain a local part and a domain, and therefore, for them, an empty
2396 address still always fails if the pattern is not itself empty.
2398 30. Exim went into a mad DNS loop when attempting to do a callout where the
2399 host was specified on an smtp transport, and looking it up yielded more
2400 than one IP address.
2402 31. Re-factored the code for checking spool and log partition space into a
2403 function that finds that data and another that does the check. The former
2404 is then used to implement four new variables: $spool_space, $log_space,
2405 $spool_inodes, and $log_inodes.
2407 32. The RFC2047 encoding function was originally intended for short strings
2408 such as real names; it was not keeping to the 75-character limit for
2409 encoded words that the RFC imposes. It now respects the limit, and
2410 generates multiple encoded words if necessary. To be on the safe side, I
2411 have increased the buffer size for the ${rfc2047: expansion operator from
2414 33. It is now permitted to omit both strings after an "if" condition; if the
2415 condition is true, the result is "true". As before, when the second string
2416 is omitted, a false condition yields an empty string. This makes it less
2417 cumbersome to write custom ACL and router conditions.
2419 34. Failure to deliver a bounce message always caused it to be frozen, even if
2420 there was an errors_to setting on the router. The errors_to setting is now
2423 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2424 canonical form (fully expanded) before being placed in
2425 $sender_host_address.
2427 36. The table in the code that translates DNS record types into text (T_A to
2428 "A" for instance) was missing entries for NS and CNAME. It is just possible
2429 that this could have caused confusion if both these types were looked up
2430 for the same domain, because the text type is used as part of Exim's
2431 per-process caching. But the chance of anyone hitting this buglet seems
2434 37. The dnsdb lookup has been extended in a number of ways.
2436 (1) There is a new type, "zns", which walks up the domain tree until it
2437 finds some nameserver records. It should be used with care.
2439 (2) There is a new type, "mxh", which is like "mx" except that it returns
2440 just the host names, not the priorities.
2442 (3) It is now possible to give a list of domains (or IP addresses) to be
2443 looked up. The behaviour when one of the lookups defers can be
2444 controlled by a keyword.
2446 (4) It is now possible to specify the separator character for use when
2447 multiple records are returned.
2449 38. The dnslists ACL condition has been extended: it is now possible to supply
2450 a list of IP addresses and/or domains to be looked up in a particular DNS
2453 39. Added log_selector=+queue_time_overall.
2455 40. When running the queue in the test harness, wait just a tad after forking a
2456 delivery process, to get repeatability of debugging output.
2458 41. Include certificate and key file names in error message when GnuTLS fails
2459 to set them up, because the GnuTLS error message doesn't include the name
2460 of the failing file when there is a problem reading it.
2462 42. Allow both -bf and -bF in the same test run.
2464 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2466 44. The "Exiscan patch" is now merged into the mainline Exim source.
2468 45. Sometimes the final signoff response after QUIT could fail to get
2469 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2470 < 0 before doing a fflush(). This bug looks as though it goes back to the
2471 introduction of TLS in release 3.20, but "sometimes" must have been rare
2472 because the tests only now provoked it.
2474 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2475 (this can affect the format of dates).
2477 47. exim_tidydb, when checking for the continued existence of a message for
2478 which it has found a message-specific retry record, was not finding
2479 messages that were in split spool directories. Consequently, it was
2480 deleting retry records that should have stayed in existence.
2482 48. Steve fixed some bugs in eximstats.
2484 49. The SPA authentication driver was not abandoning authentication and moving
2485 on to the next authenticator when an expansion was forced to fail,
2486 contradicting the general specification for all authenticators. Instead it
2487 was generating a temporary error. It now behaves as specified.
2489 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2490 (the order specifies the preference for clients). The order is now AES256,
2491 AES128, 3DES, ARCFOUR128.
2493 51. Small patch to Sieve code - explicitly set From: when generating an
2496 52. Exim crashed if a remote delivery caused a very long error message to be
2497 recorded - for instance if somebody sent an entire SpamAssassin report back
2498 as a large number of 550 error lines. This bug was coincidentally fixed by
2499 increasing the size of one of Exim's internal buffers (big_buffer) that
2500 happened as part of the Exiscan merge. However, to be on the safe side, I
2501 have made the code more robust (and fixed the comments that describe what
2504 53. Now that there can be additional text after "Completed" in log lines (if
2505 the queue_time_overall log selector is set), a one-byte patch to exigrep
2506 was needed to allow it to recognize "Completed" as not the last thing in
2509 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2510 patch that reportedly fixes this has been added. I am not expert enough to
2511 create a test for it. This is what the patch creator wrote:
2513 "I found a little strange behaviour of ldap code when working with
2514 Windows 2003 AD Domain, where users was placed in more than one
2515 Organization Units. When I tried to give exim partial DN, the exit code
2516 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2517 But simultaneously result of request was absolutely normal ldap result,
2518 so I produce this patch..."
2520 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2521 REFERENCE, so I have modified the code to exclude the patch when that macro
2524 55. Some experimental protocols are using DNS PTR records for new purposes. The
2525 keys for these records are domain names, not reversed IP addresses. The
2526 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2527 leaves it alone. Component reversal etc. now happens only for IP addresses.
2530 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2532 57. Double the size of the debug message buffer (to 2048) so that more of very
2533 long debug lines gets shown.
2535 58. The exicyclog utility now does better if the number of log files to keep
2536 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2538 59. Two changes related to the smtp_active_hostname option:
2540 (1) $smtp_active_hostname is now available as a variable.
2541 (2) The default for smtp_banner uses $smtp_active_hostname instead
2542 of $primary_hostname.
2544 60. The host_aton() function is supposed to be passed a string that is known
2545 to be a valid IP address. However, in the case of IPv6 addresses, it was
2546 not checking this. This is a hostage to fortune. Exim now panics and dies
2547 if the condition is not met. A case was found where this could be provoked
2548 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2549 components; fortuitously, this particular loophole had already been fixed
2550 by change 4.50/55 above.
2552 If there are any other similar loopholes, the new check in host_aton()
2553 itself should stop them being exploited. The report I received stated that
2554 data on the command line could provoke the exploit when Exim was running as
2555 exim, but did not say which command line option was involved. All I could
2556 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2557 running as the user.
2560 61. There was a buffer overflow vulnerability in the SPA authentication code
2561 (which came originally from the Samba project). I have added a test to the
2562 spa_base64_to_bits() function which I hope fixes it.
2565 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2566 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2568 63. The daemon start-up calls getloadavg() while still root for those OS that
2569 need the first call to be done as root, but it missed one case: when
2570 deliver_queue_load_max is set with deliver_drop_privilege. This is
2571 necessary for the benefit of the queue runner, because there is no re-exec
2572 when deliver_drop_privilege is set.
2574 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2575 This has been fixed.
2577 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2578 was in use, was not putting the data itself into the right store pool;
2579 consequently, it could be overwritten for a subsequent message in the same
2580 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2583 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2585 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2586 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2587 the function were treating the return as a boolean value, which happened to
2588 work because 0=false and not-0=true, but is not correct code.
2590 68. The host_aton() function was not handling scoped IPv6 addresses (those
2591 with, for example, "%eth0" on the end) correctly.
2593 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2594 negated items (that is, ~something) in unsigned ints. Some compilers
2595 apparently mutter when there is no cast.
2597 70. If an address verification called from an ACL failed, and did not produce a
2598 user-specific message (i.e. there was only a "system" message), nothing was
2599 put in $acl_verify_message. In this situation, it now puts the system
2602 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2603 session; change 4.31/43 added the unwanted input to the log line - except
2604 that it did not do this in the start of session case. It now does.
2606 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2607 This is wrong and can cause the other end to generate a synchronization
2608 error if it is another Exim or anything else that does the synchronization
2609 check. A QUIT command is no longer sent after a timeout.
2611 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2612 during host lookups.
2614 74. The defer_ok option of callout verification was not working if it was used
2615 when verifying addresses in header lines, that is, for this case:
2617 verify = header_sender/callout=defer_ok
2619 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2620 those file descriptors could be used for SMTP connections. If anything
2621 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2622 could be sent to the SMTP client, causing chaos. The daemon now opens
2623 stdin, stdout, and stderr to /dev/null when it puts itself into the
2626 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2627 log by default. The user can override this with suitable Perl magic.
2629 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2630 the log message when discard triggers, was not working for the DATA ACL or
2631 for the non-SMTP ACL.
2633 78. Error message wording change in sieve.c.
2635 79. If smtp_accept_max_per_host was set, the number of connections could be
2636 restricted to fewer than expected, because the daemon was trying to set up
2637 a new connection before checking whether the processes handling previous
2638 connections had finished. The check for completed processes is now done
2639 earlier. On busy systems, this bug wouldn't be noticed because something
2640 else would have woken the daemon, and it would have reaped the completed
2643 80. If a message was submitted locally by a user whose login name contained one
2644 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2645 It caused a spool format error. I have fixed the spool reading code. A
2646 related problem was that the "from" clause in the Received: line became
2647 illegal because of the space(s). It is now covered by ${quote_local_part.
2649 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2652 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2654 83. Patch from Sieve maintainer to fix unterminated string problem in
2655 "vacation" handling.
2657 84. Some minor changes to the Linux configuration files to help with other
2658 OS variants using glibc.
2660 85. One more patch for Sieve to update vacation handling to latest spec.
2663 ----------------------------------------------------
2664 See the note above about the 4.44 and 4.50 releases.
2665 ----------------------------------------------------
2671 1. Change 4.43/35 introduced a bug that caused file counts to be
2672 incorrectly computed when quota_filecount was set in an appendfile
2675 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2676 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2679 3. Give more explanation in the error message when the command for a transport
2680 filter fails to execute.
2682 4. There are several places where Exim runs a non-Exim command in a
2683 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2684 was being done only for the command run by the queryprogram router. It is
2685 now done for all such subprocesses. The other cases are: ${run, transport
2686 filters, and the commands run by the lmtp and pipe transports.
2688 5. Some older OS have a limit of 256 on the maximum number of file
2689 descriptors. Exim was using setrlimit() to set 1000 as a large value
2690 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2691 systems. I've change it so that if it can't get 1000, it tries for 256.
2693 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2694 was an oversight, and furthermore, ever since the addition of extra
2695 controls (e.g. 4.43/32), the checks on when to allow different forms of
2696 "control" were broken. There should now be diagnostics for all cases when a
2697 control that does not make sense is encountered.
2699 7. $recipients is now available in the predata ACL (oversight).
2701 8. Tidy the search cache before the fork to do a delivery from a message
2702 received from the command line. Otherwise the child will trigger a lookup
2703 failure and thereby defer the delivery if it tries to use (for example) a
2704 cached ldap connection that the parent has called unbind on.
2706 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2707 of $address_data from the recipient verification was clobbered by the
2708 sender verification.
2710 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2711 was its contents. (It was OK if the option was not defined at all.)
2713 11. A "Completed" log line is now written for messages that are removed from
2714 the spool by the -Mrm option.
2716 12. $host_address is now set to the target address during the checking of
2717 ignore_target_hosts.
2719 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2720 being passed; this would have caused $sender_host_name to have been used if
2721 matching the list had actually called for a host name (not very likely,
2722 since this list is usually IP addresses). A host name is now passed as
2725 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2726 code that specifies a non-restarting handler (typically sigaction() in
2727 modern systems) in an attempt to fix a rare and obscure crash bug.
2729 15. Narrowed the window for a race in the daemon that could cause it to ignore
2730 SIGCHLD signals. This is not a major problem, because they are used only to
2731 wake it up if nothing else does.
2733 16. A malformed maildirsize file could cause Exim to calculate negative values
2734 for the mailbox size or file count. Odd effects could occur as a result.
2735 The maildirsize information is now recalculated if the size or filecount
2738 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2739 support for a long time. Removed HAVE_SYS_VFS_H.
2741 18. Updated exipick to current release from John Jetmore.
2743 19. Allow an empty sender to be matched against a lookup in an address list.
2744 Previously the only cases considered were a regular expression, or an
2747 20. Exim went into a mad DNS lookup loop when doing a callout where the
2748 host was specified on the transport, if the DNS lookup yielded more than
2751 21. The RFC2047 encoding function was originally intended for short strings
2752 such as real names; it was not keeping to the 75-character limit for
2753 encoded words that the RFC imposes. It now respects the limit, and
2754 generates multiple encoded words if necessary. To be on the safe side, I
2755 have increased the buffer size for the ${rfc2047: expansion operator from
2758 22. Failure to deliver a bounce message always caused it to be frozen, even if
2759 there was an errors_to setting on the router. The errors_to setting is now
2762 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2763 canonical form (fully expanded) before being placed in
2764 $sender_host_address.
2766 24. Updated eximstats to version 1.33
2768 25. Include certificate and key file names in error message when GnuTLS fails
2769 to set them up, because the GnuTLS error message doesn't include the name
2770 of the failing file when there is a problem reading it.
2772 26. Expand error message when OpenSSL has problems setting up cert/key files.
2775 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2776 (this can affect the format of dates).
2778 28. exim_tidydb, when checking for the continued existence of a message for
2779 which it has found a message-specific retry record, was not finding
2780 messages that were in split spool directories. Consequently, it was
2781 deleting retry records that should have stayed in existence.
2783 29. eximstats updated to version 1.35
2784 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2785 1.35 - bugfix such that pie charts by volume are generated correctly
2787 30. The SPA authentication driver was not abandoning authentication and moving
2788 on to the next authenticator when an expansion was forced to fail,
2789 contradicting the general specification for all authenticators. Instead it
2790 was generating a temporary error. It now behaves as specified.
2792 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2793 (the order specifies the preference for clients). The order is now AES256,
2794 AES128, 3DES, ARCFOUR128.
2796 31. Small patch to Sieve code - explicitly set From: when generating an
2799 32. Exim crashed if a remote delivery caused a very long error message to be
2800 recorded - for instance if somebody sent an entire SpamAssassin report back
2801 as a large number of 550 error lines. This bug was coincidentally fixed by
2802 increasing the size of one of Exim's internal buffers (big_buffer) that
2803 happened as part of the Exiscan merge. However, to be on the safe side, I
2804 have made the code more robust (and fixed the comments that describe what
2807 33. Some experimental protocols are using DNS PTR records for new purposes. The
2808 keys for these records are domain names, not reversed IP addresses. The
2809 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2810 leaves it alone. Component reversal etc. now happens only for IP addresses.
2813 34. The host_aton() function is supposed to be passed a string that is known
2814 to be a valid IP address. However, in the case of IPv6 addresses, it was
2815 not checking this. This is a hostage to fortune. Exim now panics and dies
2816 if the condition is not met. A case was found where this could be provoked
2817 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2818 components; fortuitously, this particular loophole had already been fixed
2819 by change 4.50/55 or 4.44/33 above.
2821 If there are any other similar loopholes, the new check in host_aton()
2822 itself should stop them being exploited. The report I received stated that
2823 data on the command line could provoke the exploit when Exim was running as
2824 exim, but did not say which command line option was involved. All I could
2825 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2826 running as the user.
2829 35. There was a buffer overflow vulnerability in the SPA authentication code
2830 (which came originally from the Samba project). I have added a test to the
2831 spa_base64_to_bits() function which I hope fixes it.
2834 36. The daemon start-up calls getloadavg() while still root for those OS that
2835 need the first call to be done as root, but it missed one case: when
2836 deliver_queue_load_max is set with deliver_drop_privilege. This is
2837 necessary for the benefit of the queue runner, because there is no re-exec
2838 when deliver_drop_privilege is set.
2840 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2841 was in use, was not putting the data itself into the right store pool;
2842 consequently, it could be overwritten for a subsequent message in the same
2843 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2846 38. Sometimes the final signoff response after QUIT could fail to get
2847 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2848 < 0 before doing a fflush(). This bug looks as though it goes back to the
2849 introduction of TLS in release 3.20, but "sometimes" must have been rare
2850 because the tests only now provoked it.
2856 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2857 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2858 Now it returns an integer. A number of calls were still expecting a T/F
2859 return. Fortuitously, in all cases, the tests worked in OK situations,
2860 which is the norm. However, things would have gone wrong on any write
2861 failures on the smtp file descriptor. This function is used when sending
2862 messages over SMTP and also when doing verify callouts.
2864 2. When Exim is called to do synchronous delivery of a locally submitted
2865 message (the -odf or -odi options), it no longer closes stderr before doing
2868 3. Implemented the mua_wrapper option.
2870 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2872 5. Implemented the functions header_remove(), header_testname(),
2873 header_add_at_position(), and receive_remove_recipient(), and exported them
2876 6. If an ACL "warn" statement specified the addition of headers, Exim already
2877 inserted X-ACL-Warn: at the start if there was no header name. However, it
2878 was not making this test for the second and subsequent header lines if
2879 there were newlines in the string. This meant that an invalid header could
2880 be inserted if Exim was badly configured.
2882 7. Allow an ACL "warn" statement to add header lines at the start or after all
2883 the Received: headers, as well as at the end.
2885 8. Added the rcpt_4xx retry error code.
2887 9. Added postmaster_mailfrom=xxx to callout verification option.
2889 10. Added mailfrom=xxxx to the callout verification option, for verify=
2892 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2893 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2895 12. Inserted some casts to stop certain compilers warning when using pointer
2896 differences as field lengths or precisions in printf-type calls (mostly
2897 affecting debugging statements).
2899 13. Added optional readline() support for -be (dynamically loaded).
2901 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2902 same clock tick as a message's arrival, so that its received time was the
2903 same as the "first fail" time on the retry record, and that message
2904 remained on the queue past the ultimate address timeout, every queue runner
2905 would try a delivery (because it was past the ultimate address timeout) but
2906 after another failure, the ultimate address timeout, which should have then
2907 bounced the address, did not kick in. This was a "< instead of <=" error;
2908 in most cases the first failure would have been in the next clock tick
2909 after the received time, and all would be well.
2911 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2912 being recognized when the domain list was tested by the match_domain
2913 condition in an expansion string.
2915 16. Added the ${str2b64: operator.
2917 17. Exim was always calling setrlimit() to set a large limit for the number of
2918 processes, without checking whether the existing limit was already
2919 adequate. (It did check for the limit on file descriptors.) Furthermore,
2920 errors from getrlimit() and setrlimit() were being ignored. Now they are
2921 logged to the main and panic logs, but Exim does carry on, to try to do its
2922 job under whatever limits there are.
2924 18. Imported PCRE 5.0.
2926 19. Trivial typo in log message " temporarily refused connection" (the leading
2929 20. If the log selector return_path_on_delivery was set and an address was
2930 redirected to /dev/null, the delivery process crashed because it assumed
2931 that a return path would always be set for a "successful" delivery. In this
2932 case, the whole delivery is bypassed as an optimization, and therefore no
2935 21. Internal re-arrangement: the function for sending a challenge and reading
2936 a response while authentication was assuming a zero-terminated challenge
2937 string. It's now changed to take a pointer and a length, to allow for
2938 binary data in such strings.
2940 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2942 23. Exim was not respecting finduser_retries when seeking the login of the
2943 uid under which it was called; it was always trying 10 times. (The default
2944 setting of finduser_retries is zero.) Also, it was sleeping after the final
2945 failure, which is pointless.
2947 24. Implemented tls_on_connect_ports.
2949 25. Implemented acl_smtp_predata.
2951 26. If the domain in control=submission is set empty, Exim assumes that the
2952 authenticated id is a complete email address when it generates From: or
2953 Sender: header lines.
2955 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2956 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2957 chown and chgrp in /bin and hostname in /usr/bin.
2959 28. Exim was keeping the "process log" file open after each use, just as it
2960 does for the main log. This opens the possibility of it remaining open for
2961 long periods when the USR1 signal hits a daemon. Occasional processlog
2962 errors were reported, that could have been caused by this. Anyway, it seems
2963 much more sensible not to leave this file open at all, so that is what now
2966 29. The long-running daemon process does not normally write to the log once it
2967 has entered its main loop, and it closes the log before doing so. This is
2968 so that log files can straightforwardly be renamed and moved. However,
2969 there are a couple of unusual error situations where the daemon does write
2970 log entries, and I had neglected to close the log afterwards.
2972 30. The text of an SMTP error response that was received during a remote
2973 delivery was being truncated at 512 bytes. This is too short for some of
2974 the long messages that one sometimes sees. I've increased the limit to
2977 31. It is now possible to make retry rules that apply only when a message has a
2978 specific sender, in particular, an empty sender.
2980 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2981 it possible to be selective about when SMTP synchronization is enforced.
2983 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2985 32. Implemented hosts_connection_nolog.
2987 33. Added an ACL for QUIT.
2989 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2992 35. Added mailbox_size and mailbox_filecount to appendfile.
2994 36. Added control = no_multiline_responses to ACLs.
2996 37. There was a bug in the logic of the code that waits for the clock to tick
2997 in the case where the clock went backwards by a substantial amount such
2998 that the microsecond fraction of "now" was more than the microsecond
2999 fraction of "then" (but the whole seconds number was less).
3001 38. Added support for the libradius Radius client library this is found on
3002 FreeBSD (previously only the radiusclient library was supported).
3008 1. When certain lookups returned multiple values in the form name=value, the
3009 quoting of the values was not always being done properly. Specifically:
3010 (a) If the value started with a double quote, but contained no whitespace,
3012 (b) If the value contained whitespace other than a space character (i.e.
3013 tabs or newlines or carriage returns) it was not quoted.
3014 This fix has been applied to the mysql and pgsql lookups by writing a
3015 separate quoting function and calling it from the lookup code. The fix
3016 should probably also be applied to nisplus, ibase and oracle lookups, but
3017 since I cannot test any of those, I have not disturbed their existing code.
3019 2. A hit in the callout cache for a specific address caused a log line with no
3020 reason for rejecting RCPT. Now it says "Previous (cached) callout
3021 verification failure".
3023 3. There was an off-by-one bug in the queryprogram router. An over-long
3024 return line was truncated at 256 instead of 255 characters, thereby
3025 overflowing its buffer with the terminating zero. As well as fixing this, I
3026 have increased the buffer size to 1024 (and made a note to document this).
3028 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3029 when Exim is waiting for an SMTP response from a remote server, Exim
3030 restarts its select() call on the socket, thereby resetting its timeout.
3031 This is not a problem when such interrupts are rare. Somebody set up a cron
3032 job to run exiwhat every 2 minutes, which is less than the normal select()
3033 timeout (5 or 10 minutes). This meant that the select() timeout never
3034 kicked in because it was always reset. I have fixed this by comparing the
3035 time when an interrupt arrives with the time at the start of the first call
3036 to select(). If more time than the timeout has elapsed, the interrupt is
3037 treated as a timeout.
3039 5. Some internal re-factoring in preparation for the addition of Sieve
3040 extensions (by MH). In particular, the "personal" test is moved to a
3041 separate function, and given an option for scanning Cc: and Bcc: (which is
3042 not set for Exim filters).
3044 6. When Exim created an email address using the login of the caller as the
3045 local part (e.g. when creating a From: or Sender: header line), it was not
3046 quoting the local part when it contained special characters such as @.
3048 7. Installed new OpenBSD configuration files.
3050 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3051 try to make them clearer.
3053 9. Callout options, other than the timeout value, were being ignored when
3054 verifying sender addresses in header lines. For example, when using
3056 verify = header_sender/callout=no_cache
3058 the cache was (incorrectly) being used.
3060 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3061 only the Cygwin environment.
3063 11. When return_path_on_delivery was set as a log selector, if different remote
3064 addresses in the same message used different return paths and parallel
3065 remote delivery occurred, the wrong values would sometimes be logged.
3066 (Whenever a remote delivery process finished, the return path value from
3067 the most recently started remote delivery process was logged.)
3069 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3070 lines when AUTH and/or TLS are in use. This is the "received protocol"
3071 field. Exim used to use "asmtp" for authenticated SMTP, without any
3072 indication (in the protocol name) for TLS use. Now it follows the RFC and
3073 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3074 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3075 names appear in log lines as well as in Received: header lines.
3077 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3078 extensions, and comparison tests, and to fix some bugs.
3080 14. Changes to the "personal" filter test:
3082 (1) The test was buggy in that it was just doing the equivalent of
3083 "contains" tests on header lines. For example, if a user's address was
3084 anne@some.where, the "personal" test would incorrectly be true for
3086 To: susanne@some.where
3088 This test is now done by extracting each address from the header in turn,
3089 and checking the entire address. Other tests that are part of "personal"
3090 are now done using regular expressions (for example, to check local parts
3091 of addresses in From: header lines).
3093 (2) The list of non-personal local parts in From: addresses has been
3094 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3095 taken from the Sieve specification recommendations.
3097 (3) If the message contains any header line starting with "List-" it is
3098 treated as non-personal.
3100 (4) The test for "circular" in the Subject: header line has been removed
3101 because it now seems ill-conceived.
3103 15. Minor typos in src/EDITME comments corrected.
3105 16. Installed latest exipick from John Jetmore.
3107 17. If headers_add on a router specified a text string that was too long for
3108 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3109 of string_sprintf() is now avoided.
3111 18. $message_body_size was not set (it was always zero) when running the DATA
3112 ACL and the local_scan() function.
3114 19. For the "mail" command in an Exim filter, no default was being set for
3115 the once_repeat time, causing a random time value to be used if "once" was
3116 specified. (If the value happened to be <= 0, no repeat happened.) The
3117 default is now 0s, meaning "never repeat". The "vacation" command was OK
3118 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3119 (I found it when inspecting the code).
3121 20. There is now an overall timeout for performing a callout verification. It
3122 defaults to 4 times the callout timeout, which applies to individual SMTP
3123 commands during the callout. The overall timeout applies when there is more
3124 than one host that can be tried. The timeout is checked before trying the
3125 next host. This prevents very long delays if there are a large number of
3126 hosts and all are timing out (e.g. when the network connections are timing
3127 out). The value of the overall timeout can be changed by specifying an
3128 additional sub-option for "callout", called "maxwait". For example:
3130 verify = sender/callout=5s,maxwait=20s
3132 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3133 to the end before writing, but this should make it even safer).
3135 22. Exim was forgetting that it had advertised PIPELINING for the second and
3136 subsequent messages on an SMTP connection. It was also not resetting its
3137 memory on STARTTLS and an internal HELO.
3139 23. When Exim logs an SMTP synchronization error within a session, it now
3140 records whether PIPELINING has been advertised or not.
3142 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3143 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3144 rather than long int.
3146 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3148 26. Added the never_mail option to autoreply.
3154 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3155 crash if the getsockname() call failed; this can happen if a connection is
3156 closed very soon after it is established. The problem was simply in the
3157 order in which certain operations were done, causing Exim to try to write
3158 to the SMTP stream before it had set up the file descriptor. The bug has
3159 been fixed by making things happen in the correct order.
3165 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3166 before the connection was closed, thus losing the rejection response.
3168 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3169 some early Solaris releases, but causes trouble in current releases where
3170 socklen_t is defined.
3172 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3175 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3178 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3179 that releasing the top of it at the end releases what was used for sub-
3180 expansions (unless the block got too big). However, discard this block if
3181 the first thing is a variable or header, so that we can use its block when
3182 it is dynamic (useful for very large $message_headers, for example).
3184 6. Lookups now cache *every* query, not just the most recent. A new, separate
3185 store pool is used for this. It can be recovered when all lookup caches are
3186 flushed. Lookups now release memory at the end of their result strings.
3187 This has involved some general refactoring of the lookup sources.
3189 7. Some code has been added to the store_xxx() functions to reduce the amount
3190 of flapping under certain conditions.
3192 8. log_incoming_interface used to affect only the <= reception log lines. Now
3193 it causes the local interface and port to be added to several more SMTP log
3194 lines, for example "SMTP connection from", and rejection lines.
3196 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3198 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3200 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3201 could be overwritten at the end of the current message (or the start of a
3202 new message if it was set in a HELO ACL). The value is now preserved for
3203 the duration of the SMTP connection.
3205 12. If a transport had a headers_rewrite setting, and a matching header line
3206 contained an unqualified address, that address was qualified, even if it
3207 did not match any rewriting rules. The underlying bug was that the values
3208 of the flags that permit the existence of unqualified sender and recipient
3209 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3210 non-local messages, and by -bnq for local messages) were not being
3211 preserved with the message after it was received.
3213 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3214 "next input=" as part of the text comprising the host identity instead of
3215 the correct text. The code was using the same buffer for two different
3216 strings. However, depending on which order the printing function evaluated
3217 its arguments, the bug did not always show up. Under Linux, for example, my
3218 test suite worked just fine.
3220 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3221 doesn't work with some older versions of Perl. It has been changed to "my",
3222 which in any case is probably the better facility to use.
3224 15. A really picky compiler found some instances of statements for creating
3225 error messages that either had too many or two few arguments for the format
3228 16. The size of the buffer for calls to the DNS resolver has been increased
3229 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3230 for addresses that have a lot of PTR records. This alleviates a problem; it
3231 does not fully solve it.
3233 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3234 buffer now truncates the list and logs the incident, which is the same
3235 action as happens when Exim is looking up a host name and its aliases.
3236 Previously in this situation something unpredictable would happen;
3237 sometimes it was "internal error: store_reset failed".
3239 18. If a server dropped the connection unexpectedly when an Exim client was
3240 using GnuTLS and trying to read a response, the client delivery process
3241 crashed while trying to generate an error log message.
3243 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3244 string, for example:
3246 warn message = H1: something\nH2: something
3248 the text was added as a single header line from Exim's point of view
3249 though it ended up OK in the delivered message. However, searching for the
3250 second and subsequent header lines using $h_h2: did not work. This has been
3251 fixed. Similarly, if a system filter added multiple headers in this way,
3252 the routers could not see them.
3254 20. Expanded the error message when iplsearch is called with an invalid key to
3255 suggest using net-iplsearch in a host list.
3257 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3258 ACLs are no longer actually imposed (and a message to that effect is
3261 22. If a "gecos" field in a passwd entry contained escaped characters, in
3262 particular, if it contained a \" sequence, Exim got it wrong when building
3263 a From: or a Sender: header from that name. A second bug also caused
3264 incorrect handling when an unquoted " was present following a character
3265 that needed quoting.
3267 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3268 was not being matched caselessly.
3270 24. Arranged for all hyphens in the exim.8 source to be escaped with
3273 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3274 properly. Recipient callout cache records were still being keyed to include
3275 the sender, even when use_sender was set false. This led to far more
3276 callouts that were necessary. The sender is no longer included in the key
3277 when use_sender is false.
3279 26. Added "control = submission" modifier to ACLs.
3281 27. Added the ${base62d: operator to decode base 62 numbers.
3283 28. dnsdb lookups can now access SRV records.
3285 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3286 the configuration file.
3288 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3289 (-v) mode. This makes the output for a verbose queue run more intelligible.
3291 31. Added a use_postmaster feature to recipient callouts.
3293 32. Added the $body_zerocount variable, containing the number of binary zero
3294 bytes in the message body.
3296 33. The time of last modification of the "new" subdirectory is now used as the
3297 "mailbox time last read" when there is a quota error for a maildir
3300 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3302 35. Added +ignore_unknown as a special item in host lists.
3304 36. Code for decoding IPv6 addresses in host lists is now included, even if
3305 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3306 address was recognized as an IP address, but was then not correctly decoded
3307 into binary, causing unexpected and incorrect effects when compared with
3314 1. Very minor rewording of debugging text in manualroute to say "list of
3315 hosts" instead of "hostlist".
3317 2. If verify=header_syntax was set, and a header line with an unqualified
3318 address (no domain) and a large number of spaces between the end of the
3319 name and the colon was received, the reception process suffered a buffer
3320 overflow, and (when I tested it) crashed. This was caused by some obsolete
3321 code that should have been removed. The fix is to remove it!
3323 3. When running in the test harness, delay a bit after writing a bounce
3324 message to get a bit more predictability in the log output.
3326 4. Added a call to search_tidyup() just before forking a reception process. In
3327 theory, someone could use a lookup in the expansion of smtp_accept_max_
3328 per_host which, without the tidyup, could leave open a database connection.
3330 5. Added the variables $recipient_data and $sender_data which get set from a
3331 lookup success in an ACL "recipients" or "senders" condition, or a router
3332 "senders" option, similar to $domain_data and $local_part_data.
3334 6. Moved the writing of debug_print from before to after the "senders" test
3337 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3338 problems for message scanning, either using a data ACL, or using
3339 local_scan() because the Received: header was not generated till after they
3340 were called (in order to set the time as the time of reception completion).
3341 I have revised the way this works. The header is now generated after the
3342 body is received, but before the ACL or local_scan() are called. After they
3343 are run, the timestamp in the header is updated.
3349 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3350 before starting a queue runner without re-exec. This happened only when
3351 deliver_drop_privilege was set or when the Exim user was set to root. The
3352 effect of the bug was that timeouts during subsequent deliveries caused
3353 crashes instead of being properly handled. The handler is now left at its
3354 default (and expected) setting.
3356 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3357 message, again when deliver_drop_privilege is set or Exim is run as root.
3358 The bug described in (1) was not present in this case, but the tidying up
3359 of the other signals was missing. I have made the two cases consistent.
3361 3. The ignore_target_hosts setting on a manualroute router was being ignored
3362 for hosts that were looked up using the /MX notation.
3364 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3367 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3368 operated on the sender address. After changing the $sender_address to <>
3369 for the sender address verify, Exim was re-instated it as the original
3370 (before rewriting) address, but remembering that it had rewritten it, so it
3371 wasn't rewriting it again. This bug also had the effect of breaking the
3372 sender address verification caching when the sender address was rewritten.
3374 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3375 This has been changed so that if the ip literal address matches
3376 ignore_target_hosts, the router declines.
3378 7. Added expansion conditions match_domain, match_address, and match_local_
3379 part (NOT match_host).
3381 8. The placeholder for the Received: header didn't have a length field set.
3383 9. Added code to Exim itself and to exim_lock to test for a specific race
3384 condition that could lead to file corruption when using MBX delivery. The
3385 issue is with the lockfile that is created in /tmp. If this file is removed
3386 after a process has opened it but before that process has acquired a lock,
3387 there is the potential for a second process to recreate the file and also
3388 acquire a lock. This could lead to two Exim processes writing to the file
3389 at the same time. The added code performs the same test as UW imapd; it
3390 checks after acquiring the lock that its file descriptor still refers to
3391 the same named file.
3393 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3394 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3397 11. Added the smtp_active_hostname option. If used, this will typically be made
3398 to depend on the incoming interface address. Because $interface_address is
3399 not set up until the daemon has forked a reception process, error responses
3400 that can happen earlier (such as "too many connections") no longer contain
3403 12. If an expansion in a condition on a "warn" statement fails because a lookup
3404 defers, the "warn" statement is abandoned, and the next ACL statement is
3405 processed. Previously this caused the whole ACL to be aborted.
3407 13. Added the iplsearch lookup type.
3409 14. Added ident_timeout as a log selector.
3411 15. Added tls_certificate_verified as a log selector.
3413 16. Added a global option tls_require_ciphers (compare the smtp transport
3414 option of the same name). This controls incoming TLS connections.
3416 17. I finally figured out how to make tls_require_ciphers do a similar thing
3417 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3418 before starting the TLS session.
3420 18. Tabs are now shown as \t in -bP output.
3422 19. If the log selector return_path_on_delivery was set, Exim crashed when
3423 bouncing a message because it had too many Received: header lines.
3425 20. If two routers both had headers_remove settings, and the first one included
3426 a superfluous trailing colon, the final name in the first list and the
3427 first name in the second list were incorrectly joined into one item (with a
3428 colon in the middle).
3434 1. Added -C and -D options to the exinext utility, mainly to make it easier
3435 to include in the automated testing, but these could be helpful when
3436 multiple configurations are in use.
3438 2. The exinext utility was not formatting the output nicely when there was
3439 an alternate port involved in the retry record key, nor when there was a
3440 message id as well (for retries that were specific to a specific message
3441 and a specific host). It was also confused by IPv6 addresses, because of
3442 the additional colons they contain. I have fixed the IPv4 problem, and
3443 patched it up to do a reasonable job for IPv6.
3445 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3446 delivery, the log line now contains "pipelined" if PIPELINING was used.
3448 4. An SMTP transport process used to panic and die if the bind() call to set
3449 an explicit outgoing interface failed. This has been changed; it is now
3450 treated in the same way as a connect() failure.
3452 5. A reference to $sender_host_name in the part of a conditional expansion
3453 that was being skipped was still causing a DNS lookup. This no longer
3456 6. The def: expansion condition was not recognizing references to header lines
3457 that used bh_ and bheader_.
3459 7. Added the _cache feature to named lists.
3461 8. The code for checking quota_filecount in the appendfile transport was
3462 allowing one more file than it should have been.
3464 9. For compatibility with Sendmail, the command line option
3472 and sets the incoming protocol and host name (for trusted callers). The
3473 host name and its colon can be omitted when only the protocol is to be set.
3474 Note the Exim already has two private options, -pd and -ps, that refer to
3475 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3476 "s", but I don't think that's a major issue.
3478 10. A number of refactoring changes to the code, none of which should affect
3481 (a) The number of logging options was getting close to filling up the
3482 32-bit word that was used as a bit map. I have split them into two classes:
3483 those that are passed in the argument to log_write(), and those that are
3484 only ever tested independently outside of that function. These are now in
3485 separate 32-bit words, so there is plenty of room for expansion again.
3486 There is no change in the user interface or the logging behaviour.
3488 (b) When building, for example, log lines, the code previously used a
3489 macro that called string_cat() twice, in order to add two strings. This is
3490 not really sufficiently general. Furthermore, there was one instance where
3491 it was actually wrong because one of the argument was used twice, and in
3492 one call a function was used. (As it happened, calling the function twice
3493 did not affect the overall behaviour.) The macro has been replaced by a
3494 function that can join an arbitrary number of extra strings onto a growing
3497 (c) The code for expansion conditions now uses a table and a binary chop
3498 instead of a serial search (which was left over from when there were very
3499 few conditions). Also, it now recognizes conditions like "pam" even when
3500 the relevant support is not compiled in: a suitably worded error message is
3501 given if an attempt is made to use such a condition.
3503 11. Added ${time_interval:xxxxx}.
3505 12. A bug was causing one of the ddress fields not to be passed back correctly
3506 from remote delivery subprocesses. The field in question was not being
3507 subsequently used, so this caused to problems in practice.
3509 13. Added new log selectors queue_time and deliver_time.
3511 14. Might have fixed a bug in maildirsizefile handling that threw up
3512 "unexpected character" debug warnings, and recalculated the data
3513 unnecessarily. In any case, I expanded the warning message to give more
3516 15. Added the message "Restricted characters in address" to the statements in
3517 the default ACL that block characters like @ and % in local parts.
3519 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3520 Three changes have been made:
3522 (a) There was a serious bug; a negative response to MAIL caused the whole
3523 recipient domain to be cached as invalid, thereby blocking all messages
3524 to all local parts at the same domain, from all senders. This bug has
3525 been fixed. The domain is no longer cached after a negative response to
3526 MAIL if the sender used is not empty.
3528 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3531 (c) A new callout option, "use_sender" has been added for people who want
3532 the modified behaviour.
3538 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3541 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3542 indeed breaks things for older releases.
3544 3. Added additional logging to the case where there is a problem reading data
3545 from a filter that is running in a subprocess using a pipe, in order to
3546 try to track down a specific problem.
3548 4. Testing facility fudge: when running in the test harness and attempting
3549 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3550 getting "No route to host". Convert this to a timeout.
3552 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3555 6. Some OS don't have socklen_t but use size_t instead. This affects the
3556 fifth argument of getsockopt() amongst other things. This is now
3557 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3558 can be set for individual OS. I have set it for SunOS5, OSF1, and
3559 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3560 some earlier ones do not.
3562 7. Change 4.30/15 was not doing the test caselessly.
3564 8. The standard form for an IPv6 address literal was being rejected by address
3565 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3566 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3567 this, as well as the form without the "IPv6" on the front (but only when
3568 address literals are enabled, of course).
3570 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3572 10. Exim crashed if a message with an empty sender address specified by -f
3573 encountered a router with an errors_to setting. This could be provoked only
3574 by a command such as
3578 where an empty string was supplied; "<>" did not hit this bug.
3580 11. Installed PCRE release 4.5.
3582 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3583 remained set. It is now erased.
3585 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3586 times from message ids (which are base 36 rather than the normal 62).
3588 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3589 were being counted as actual protocol errors, and logged if the log
3590 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3591 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3592 and DATA following a set of rejected RCPTs do not count as protocol errors.
3593 In other words, Exim assumes they were pipelined, though this may not
3594 actually be the case. Of course, in all cases the client gets an
3595 appropriate error code.
3597 15. If a lookup fails in an ACL condition, a message about the failure may
3598 be available; it is used if testing the ACL cannot continue, because most
3599 such messages specify what the cause of the deferral is. However, some
3600 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3601 that caused an old message to be retained and used if a later statement
3602 caused a defer, replacing the real cause of the deferral.
3604 16. If an IP address had so many PTR records that the DNS lookup buffer
3605 was not large enough to hold them, Exim could crash while trying to process
3606 the truncated data. It now detects and logs this case.
3608 17. Further to 4.21/58, another change has been made: if (and only if) the
3609 first line of a message (the first header line) ends with CRLF, a bare LF
3610 in a subsequent header line has a space inserted after it, so as not to
3611 terminate the header.
3613 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3614 unnecessarily, used atoi() instead of strtol(), and didn't check the
3615 termination when getting file sizes from file names by regex.
3617 19. Completely re-implemented the support for maildirsize files, in the light
3618 of a number of problems with the previous contributed implementation
3619 (4.30/29). In particular:
3621 . If the quota is zero, the maildirsize file is maintained, but no quota is
3624 . If the maildir directory does not exist, it is created before any attempt
3625 to write a maildirsize file.
3627 . The quota value in the file is just a cache; if the quota is changed in
3628 the transport, the new value overrides.
3630 . A regular expression is available for excluding directories from the
3633 20. The autoreply transport checks the characters in options that define the
3634 message's headers; it allows continued headers, but it was checking with
3635 isspace() after an embedded newline instead of explicitly looking for a
3638 21. If all the "regular" hosts to which an address was routed had passed their
3639 expiry times, and had not reached their retry times, the address was
3640 bounced, even if fallback hosts were defined. Now Exim should go on to try
3643 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3644 equivalent code in the SMTP transport. Some hosts send humungous responses
3645 to HELO/EHLO, more than 1024 it seems.
3647 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3648 gives compiler warnings in some environments. I've now done it "properly",
3651 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3652 (because of problems with the built-in one) was declared to return uschar *
3653 instead of char *, causing compiler failure.
3655 25. Fixed a file descriptor leak when processing alias/forward files.
3657 26. Fixed a minor format string issue in dbfn.c.
3659 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3661 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3662 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3665 29. When a sender address is verified, it is cached, to save repeating the test
3666 when there is more than one recipient in a message. However, when the
3667 verification involves a callout, it is possible for different callout
3668 options to be set for different recipients. It is too complicated to keep
3669 track of this in the cache, so now Exim always runs a verification when a
3670 callout is required, relying on the callout cache for the optimization.
3671 The overhead is duplication of the address routing, but this should not be
3674 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3675 to be verified with callout=postmaster, and the main callout worked but the
3676 postmaster check failed, the verification correctly failed. However, if a
3677 subsequent RCPT command asked for sender verification *without* the
3678 postmaster check, incorrect caching caused this verification also to fail,
3681 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3682 it was not caching the DNS options (qualify_single, search_parents) that
3683 were used when the lookup failed. A subsequent lookup with different
3684 options therefore always gave the same answer, though there were cases
3685 where it should not have. (Example: a "domains = !$mx_any" option on a
3686 dnslookup router: the "domains" option is always processed without any
3687 widening, but the router might have qualify_single set.) Now Exim uses the
3688 cached value only when the same options are set.
3690 32. Added John Jetmore's "exipick" utility to the distribution.
3692 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3693 than a timeout (e.g. a certificate is required, and is not provided), an
3694 Exim server now closes the connection immediately. Previously it waited for
3695 the client to close - but if the client is SSL, it seems that they each
3696 wait for each other, leading to a delay before one of them times out.
3698 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3699 maintained 0.8.x compatibility because I don't think many are using it, and
3700 it is clearly obsolete.
3702 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3705 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3706 client certificate was expired. A simple patch fixes this, though I don't
3707 understand the full logic of why the verify callback is called multiple
3710 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3711 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3712 which causes problems with some clients (such as the Certicom SSL Plus
3713 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3714 disables the coutermeasure allowing Eudora to connect."
3716 38. Exim was not checking that a write() to a log file succeeded. This could
3717 lead to Bad Things if a log got too big, in particular if it hit a file
3718 size limit. Exim now panics and dies if it cannot write to a log file, just
3719 as it does if it cannot open a log file.
3721 39. Modified OS/Makefile-Linux so that it now contains
3723 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3725 The two -D definitions ensure that Exim is compiled with large file
3726 support, which makes it possible to handle log files that are bigger than
3729 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3730 instance) a domain was checked against a named list that involved a lookup,
3731 causing $domain_data to be set, then another domain was checked against the
3732 same list, then the first domain was re-checked, the value of $domain_data
3733 after the final check could be wrong. In particular, if the second check
3734 failed, it could be set empty. This bug probably also applied to
3737 41. The strip_trailing_dot option was not being applied to the address given
3738 with the -f command-line option.
3740 42. The code for reading a message's header from the spool was incrementing
3741 $received_count, but never initializing it. This meant that the value was
3742 incorrect (doubled) while delivering a message in the same process in which
3743 it was received. In the most common configuration of Exim, this never
3744 happens - a fresh exec is done - but it can happen when
3745 deliver_drop_privilege is set.
3747 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3748 it now includes up to 150 characters of the unexpected data in the log
3751 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3752 and building data strings. The size of both of these buffers was 10 000
3753 bytes - far larger than anybody would *ever* want, thought I. Needless to
3754 say, somebody hit the limit. I have increased the maximum line length to
3755 20 000 and the maximum data length of concatenated lines to 100 000. I have
3756 also fixed two bugs, because there was no checking on these buffers. Tsk,
3757 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3758 buffer is too small.
3760 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3761 lsearch lookups. Now it does.
3763 46. When parsing a route_list item in a manualroute router, a fixed-length
3764 buffer was used for the list of hosts. I made this 1024 bytes long,
3765 thinking that nobody would ever have a list of hosts that long. Wrong.
3766 Somebody had a whole pile of complicated expansion conditions, and the
3767 string was silently truncated, leading to an expansion error. It turns out
3768 that it is easier to change to an unlimited length (owing to other changes
3769 that have happened since this code was originally written) than to build
3770 structure for giving a limitation error. The length of the item that
3771 expands into the list of hosts is now unlimited.
3773 47. The lsearch lookup could not handle data where the length of text line was
3774 more than 4095 characters. Such lines were truncated, leading to shortened
3775 data being returned. It should now handle lines of any length.
3777 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3778 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3781 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3782 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3783 no longer happens in comment lines. A list of replacements is now placed
3784 at the head of all of the source files, except those whose only change is
3785 to replace PERL_COMMAND in the very first #! line.
3787 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3788 messages on the queue, with a bottom-up merge sort, using code contributed
3789 by Michael Haardt. This should make operations like -bp somewhat faster on
3790 large queues. It won't affect queue runners, except when queue_run_in_order
3793 51. Installed eximstats 1.31 in the distribution.
3795 52. Added support for SRV lookups to the dnslookup router.
3797 53. If an ACL referred to $message_body or $message_body_end, the value was not
3798 reset for any messages that followed in the same SMTP session.
3800 54. The store-handling optimization for building very long strings was not
3801 differentiating between the different store pools. I don't think this
3802 actually made any difference in practice, but I've tidied it.
3804 55. While running the routers to verify a sender address, $sender_address
3805 was still set to the sender address. This is wrong, because when routing to
3806 send a bounce to the sender, it would be empty. Therefore, I have changed
3807 it so that, while verifying a sender address, $sender_address is set to <>.
3808 (There is no change to what happens when verifying a recipient address.)
3810 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3811 target A or AAAA records (if not already returned) without resetting the
3812 qualify_single or search_parents options of the DNS resolver. These are
3813 inappropriate in this case because the targets of MX and SRV records must
3814 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3815 target that, when qualified, matched something in the local domain. These
3816 two options are now turned off when doing these lookups.
3818 57. It seems that at least some releases of Reiserfs (which does not have the
3819 concept of a fixed number of inodes) returns zero and not -1 for the
3820 number of available inodes. This interacted badly with check_spool_inodes,
3821 which assumed that -1 was the "no such thing" setting. What I have done is
3822 to check that the total number of inodes is greater than zero before doing
3823 the test of how many are available.
3825 58. When a "warn" ACL statement has a log_message modifier, the message is
3826 remembered, and not repeated. This is to avoid a lot of repetition when a
3827 message has many recipients that cause the same warning to be written.
3828 Howewer, Exim was preserving the list of already written lines for an
3829 entire SMTP session, which doesn't seem right. The memory is now reset if a
3830 new message is started.
3832 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3833 debugging output unless log_rewrite was also set.
3835 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3836 of (int)(handle) when we know that handle contains (void *)(-1).
3838 61. The Exim daemon panic-logs an error return when it closes the incoming
3839 connection. However "connection reset by peer" seems to be common, and
3840 isn't really an error worthy of noting specially, so that particular error
3843 62. When Exim is trying to find all the local interfaces, it used to panic and
3844 die if the ioctl to get the interface flags failed. However, it seems that
3845 on at least one OS (Solaris 9) it is possible to have an interface that is
3846 included in the list of interfaces, but for which you get a failure error
3847 for this call. This happens when the interface is not "plumbed" into a
3848 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3849 failure of the "get flags" call assumes that the interface is down.
3851 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3852 makes life easier for people who are doing arithmetic on fields extracted
3853 from dates, where you often get leading zeros that should not be
3854 interpreted as octal.
3856 64. Added qualify_domain to the redirect router, to override the global
3859 65. If a pathologically long header line contained very many addresses (the
3860 report of this problem mentioned 10 000) and each of them was rewritten,
3861 Exim could use up a very large amount of memory. (It kept on making new
3862 copies of the header line as it rewrote, and never released the old ones.)
3863 At the expense of a bit more processing, the header rewriting function has
3864 been changed so that it no longer eats memory in this way.
3866 66. The generation of the Received: header has been moved from the time that a
3867 message starts to be received, to the time that it finishes. The timestamp
3868 in the Received: header should now be very close to that of the <= log
3869 line. There are two side-effects of this change:
3871 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3872 logged header lines no longer include the local Received: line, because
3873 it has not yet been created. The same applies to a copy of the message
3874 that is returned to a non-SMTP sender when a message is rejected.
3876 (b) When a filter file is tested using -bf, no additional Received: header
3877 is added to the test message. After some thought, I decided that this
3880 This change does not affect the value of $received_for. It is still set
3881 after address rewriting, but before local_scan() is called.
3883 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3885 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3886 gave an unhelpful panic error message, and a defer error. I have managed to
3887 change this behaviour so that it now rejects any supplied certificate,
3888 which seems right, as the list of acceptable certificates is empty.
3890 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3891 gave an unhelpful defer error. I have not managed to make this reject any
3892 supplied certificates, but the error message it gives is "no certificate
3893 supplied", which is not helpful.
3895 70. exigrep's output now also includes lines that are not associated with any
3896 message, but which match the given pattern. Implemented by a patch from
3897 Martin Sluka, which also tidied up the Perl a bit.
3899 71. Recipient callout verification, like sender verification, was using <> in
3900 the MAIL FROM command. This isn't really the right thing, since the actual
3901 sender may affect whether the remote host accepts the recipient or not. I
3902 have changed it to use the actual sender in the callout; this means that
3903 the cache record is now keyed on a recipient/sender pair, not just the
3904 recipient address. There doesn't seem to be a real danger of callout loops,
3905 since a callout by the remote host to check the sender would use <>.
3906 [SEE ABOVE: changed after hitting problems.]
3908 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3909 temporary errors. However, in the case of such a code being given after
3910 the end of a data transmission (i.e. after ".") Exim was failing to write
3911 a retry record for the message. (Yes, there was some broken host that was
3912 actually sending 8xx at this point.)
3914 73. An unknown lookup type in a host list could cause Exim to panic-die when
3915 the list was checked. (An example that provoked this was putting <; in the
3916 middle of a list instead of at the start.) If this happened during a DATA
3917 ACL check, a -D file could be left lying around. This kind of configuration
3918 error no longer causes Exim to die; instead it causes a defer errror. The
3919 incident is still logged to the main and panic logs.
3921 74. Buglet left over from Exim 3 conversion. The message "too many messages
3922 in one connection" was written to the rejectlog but not the mainlog, except
3923 when address rewriting (yes!) was being logged.
3925 75. Added write_rejectlog option.
3927 76. When a system filter was run not as root (that is, when system_filter_user
3928 was set), the values of the $n variables were not being returned to the
3929 main process; thus, they were not subsequently available in the $sn
3932 77. Added +return_path_on_delivery log selector.
3934 78. A connection timeout was being treated differently from recipients deferred
3935 when testing hosts_max_try with a message that was older than the host's
3936 retry timeout. (The host should not be counted, thus allowing all hosts to
3937 be tried at least once before bouncing.) This may have been the cause of an
3938 occasionally reported bug whereby a message would remain on the queue
3939 longer than the retry timeout, but would be bounced if a delivery was
3940 forced. I say "may" because I never totally pinned down the problem;
3941 setting up timeout/retry tests is difficult. See also the next item.
3943 79. The ultimate address timeout was not being applied to errors that involved
3944 a combination of host plus message (for example, a timeout on a MAIL
3945 command). When an address resolved to a number of possible hosts, and they
3946 were not all tried for each delivery (e.g. because of hosts_max_try), a
3947 message could remain on the queue longer than the retry timeout.
3949 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3952 81. Fixed an obscure SMTP outgoing bug which required at least the following
3953 conditions: (a) there was another message waiting for the same server;
3954 (b) the server returned 5xx to all RCPT commands in the first message so
3955 that the message was not completed; (c) the server dropped the connection
3956 or gave a negative response to the RSET that Exim sends to abort the
3957 transaction. The observed case was a dropped connection after DATA that had
3958 been sent in pipelining mode. That is, the server had advertised PIPELINING
3959 but was not implementing it correctly. The effect of the bug was incorrect
3960 behaviour, such as trying another host, and this could lead to a crash.
3966 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3967 and daemon.c were passed as pointers to ints; they should have been
3968 pointers to socklen_t variables (which are typically unsigned ints).
3970 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3973 3. Fixed a really odd bug that affected only the testing scheme; patching a
3974 certain fixed string in the binary changed the value of another string that
3975 happened to be identical to the end of the original first string.
3977 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3978 name", it returns that address as the IP address. On some operating
3979 systems (e.g. Solaris), it also passes back the IP address string as the
3980 "host name". However, on others (e.g. Linux), it passes back an empty
3981 string. Exim wasn't checking for this, and was changing the host name to an
3982 empty string, assuming it had been canonicized.
3984 5. Although rare, it is permitted to have more than one PTR record for a given
3985 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3986 all the names associated with an address, because they do in Solaris.
3987 However, it seems that they do not in Linux for data that comes from the
3988 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3989 I found this out when I moved to a new Linux workstation and tried to run
3990 the Exim test suite.
3992 To get round this problem I have changed the code so that it now does its
3993 own call to the DNS to look up PTR records when searching for a host name.
3994 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3995 addresses that are only in /etc/hosts are still found.
3997 This behaviour is, however, controlled by an option called host_lookup_
3998 order, which defaults to "bydns:byaddr". If people want to use the other
3999 order, or indeed, just use one or the other means of lookup, they can
4000 specify it in this variable.
4002 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4003 some operating systems, this comes back from gethostbyaddr() as an empty
4004 string, and this is what Exim used to test for. However, it seems that in
4005 other systems, "." is yielded. Exim now tests for this case too.
4007 7. The values of check_spool_space and check_log_space are now held internally
4008 as a number of kilobytes instead of an absolute number of bytes. If a
4009 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4010 kilobyte. This means that much larger values can be stored.
4012 8. Exim monitor: an attempt to get the action menu when not actually pointing
4013 at a message produces an empty menu entitled "No message selected". This
4014 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4015 no entries in it ("Shell widget menu has zero width and/or height"). So I
4016 have added a single, blank menu entry in this case.
4018 9. Added ${quote_local_part.
4020 10. MIME decoding is now applied to the contents of Subject: header lines when
4023 11. Now that a reference to $sender_host_address automatically causes a reverse
4024 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4025 host lookup before query-style lookups in lists that might use this
4026 variable. This has therefore been abolished, and the "net-" prefix is no
4027 longer necessary for query-style lookups.
4029 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4030 have been a typo for LFLAGS, so it has been changed.
4032 13. The install script calls Exim with "-C /dev/null" in order to find the
4033 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4034 to be output. Howeve, since Exim outputs its version number before the
4035 error, it didn't break the script. It just looked ugly. I fixed this by
4036 always allowing "-C /dev/null" if the caller is root.
4038 14. Ignore overlarge ACL variable number when reading spool file - insurance
4039 against a later release with more variables having written the file.
4041 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4042 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4043 this, as well as the form without the "IPv6" on the front.
4045 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4046 OS/Makefile-Darwin file.
4048 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4049 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4051 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4052 variable. However, some OS use ints for inodes. I've added cast to long int
4053 to get rid of the compiler warning.
4055 19. I had forgotten to lock out "/../" in configuration file names when
4056 ALT_CONFIG_PREFIX was set.
4058 20. Routers used for verification do not need to specify transports. However,
4059 if such a router generated a host list, and callout was configured, Exim
4060 crashed, because it could not find a port number from the (non-existent)
4061 transport. It now assumes port 25 in this circumstance.
4063 21. Added the -t option to exigrep.
4065 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4066 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4067 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4068 them is set, LOOKUP_LSEARCH is forced.
4070 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4072 24. Added sender and host information to the "rejected by local_scan()" log
4073 line; previously there was no indication of these.
4075 25. Added .include_if_exists.
4077 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4078 Linux. It turns out that not all file systems support this. Apparently some
4079 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4080 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4081 not supported on the file descriptor, is now ignored when Exim is trying to
4082 sync a directory. This applies only to Linux.
4084 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4086 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4087 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4090 29. Added support for maildirsize files from supplied patch (modified a bit).
4092 30. The use of :fail: followed by an empty string could lead Exim to respond to
4093 sender verification failures with (e.g.):
4095 550 Verification failed for <xxx>
4096 550 Sender verify failed
4098 where the first response line was missing the '-' that indicates it is not
4099 the final line of the response.
4101 31. The loop for finding the name of the user that called Exim had a hardwired
4102 limit of 10; it now uses the value of finduser_retries, which is used for
4103 all other user lookups.
4105 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4108 33. Exim was neglecting to zero errno before one call of strtol() when
4109 expanding a string and expecting an integer value. On some systems this
4110 resulted in spurious "integer overflow" errors. Also, it was casting the
4111 result into an int without checking.
4113 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4114 did not work. The code looks as if it has *never* worked, though it appears
4115 to have been documented since at least releast 1.62. I have made it work.
4117 35. The "timeout_DNS" error in retry rules, also documented since at least
4118 1.62, also never worked. As it isn't clear exactly what this means, and
4119 clearly it isn't a major issue, I have abolished the feature by treating it
4120 as "timeout", and writing a warning to the main and panic logs.
4122 36. The display of retry rules for -brt wasn't always showing the error code
4125 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4126 timeout_connect_A, timeout_connect_MX.
4128 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4129 to the empty sender.
4131 39. The daemon was not analysing the content of -oX till after it had closed
4132 stderr and disconnected from the controlling terminal. This meant that any
4133 syntax errors were only noted on the panic log, and the return code from
4134 the command was 0. By re-arranging the code a little, I've made the
4135 decoding happen first, so such errors now appear on stderr, and the return
4136 code is 1. However, the actual setting up of the sockets still happens in
4137 the disconnected process, so errors there are still only recorded on the
4140 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4141 connections (as happens on some IP stacks) was logged at start up time as
4142 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4143 it from "IPv6 and IPv4", which means that two separate sockets are being
4146 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4147 says whether AF_INET or AF_INET6 was passed as an argument.
4149 42. Exiwhat output was messed up when time zones were included in log
4152 43. Exiwhat now gives more information about the daemon's listening ports,
4153 and whether -tls-on-connect was used.
4155 44. The "port" option of the smtp transport is now expanded.
4157 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4158 silently ignored. Now an error message is written to the main and panic
4161 46. There's a new ACL modifier called "logwrite" which writes to a log file
4162 as soon as it is encountered.
4164 47. Added $local_user_uid and $local_user_gid at routing time.
4166 48. Exim crashed when trying to verify a sender address that was being
4169 49. Exim was recognizing only a space character after ".include". It now also
4170 recognizes a tab character.
4172 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4173 extracting the relevant information from the specification. The man page no
4174 longer contains scrambled data for the -d option, and I've added a section
4175 at the front about calling Exim under different names.
4177 51. Added "extra_headers" argument to the "mail" command in filter files.
4179 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4182 53. Installed eximstats 1.29.
4184 54. Added transport_filter_timeout as a generic transport option.
4186 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4187 Cc: header lines. This was required by RFC 822, but it not required by RFC
4190 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4191 incoming messages that did not have them. Now it does so only if the
4192 message originates locally, that is, if there is no associated remote host
4193 address. When Resent- header lines are present, this applies to the Resent-
4194 lines rather than the non-Resent- lines.
4196 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4197 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4199 58. Messages for configuration errors now include the name of the main
4200 configuration file - useful now that there may be more than one file in a
4201 list (.included file names were always shown).
4203 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4204 for those rare installations that do not start the daemon as root or run it
4205 setuid root. I've cut out the call to initgroups() if the daemon is not
4208 60. The Exim user and group can now be bound into the binary as text strings
4209 that are looked up at the start of Exim's processing.
4211 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4213 62. Added $mailstore_basename variable.
4215 63. Installed patch to sieve.c from Michael Haardt.
4217 64. When Exim failed to open the panic log after failing to open the main log,
4218 the original message it was trying to log was written to stderr and debug
4219 output, but if they were not available (the usual case in production), it
4220 was lost. Now it is written to syslog before the two lines that record the
4221 failures to open the logs.
4223 65. Users' Exim filters run in subprocesses under the user's uid. It is
4224 possible for a "deliver" command or an alias in a "personal" command to
4225 provoke an address rewrite. If logging of address rewriting is configured,
4226 this fails because the process is not running as root or exim. There may be
4227 a better way of dealing with this, but for the moment (because 4.30 needs
4228 to be released), I have disabled address rewrite logging when running a
4229 filter in a non-root, non-exim process.
4235 1. The buildconfig auxiliary program wasn't quoting the value set for
4236 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4237 not defined. This bug was masked in 4.22 by the effect that was fixed in
4240 2. Some messages that were rejected after a message id was allocated were
4241 shown as "incomplete" by exigrep. It no longer does this for messages that
4242 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4244 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4245 have allow_domain_literals set, the ID did not get logged in the <= line.
4246 Domain literals are now always recognized in Message-ID: header lines.
4248 4. The first argument for a ${extract expansion item is the key name or field
4249 number. Leading and trailing spaces in this item were not being ignored,
4250 causing some misleading effects.
4252 5. When deliver_drop_privilege was set, single queue runner processes started
4253 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4254 same command in the process it spins off) were not dropping privilege.
4256 6. When the daemon running as "exim" started a queue runner, it always
4257 re-executed Exim in the spun-off process. This is a waste of effort when
4258 deliver_drop_privilege is set. The new process now just calls the
4259 queue-runner function directly.
4265 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4268 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4269 ignored. Though the use of -oP was forcing the writing of a pid file, it
4270 was always written to the default place.
4272 3. If the message "no IP address found for host xxxx" is generated during
4273 incoming verification, it is now followed by identification of the incoming
4274 connection (so you can more easily find what provoked it).
4276 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4278 5. Added some features to "harden" Exim a bit more against certain attacks:
4280 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4281 be put in Local/Makefile. This is like the never_users runtime option,
4282 but it cannot be overridden. The default setting is "root".
4284 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4285 prefix string with which any file named in a -C command line option
4288 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4289 is retained for -C and -D only if the caller of Exim is root. Without
4290 it, the exim user may also use -C and -D and retain privilege.
4292 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4293 command line option is disabled.
4295 6. Macro names set by the -D option must start with an upper case letter, just
4296 like macro names defined in the configuration file.
4298 7. Added "dereference=" facility to LDAP.
4300 8. Two instances of the typo "uknown" in the source files are fixed.
4302 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4303 the Configure-Makefile script screwed up while processing it.
4305 10. Incorporated PCRE 4.4.
4307 11. The SMTP synchronization check was not operating right at the start of an
4308 SMTP session. For example, it could not catch a HELO sent before the client
4309 waited for the greeting. There is now a check for outstanding input at the
4310 point when the greeting is written. Because of the duplex, asynchronous
4311 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4312 way, but not yet received, when the check is performed.
4314 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4315 on TCP/IP sockets, because this apparently causes some broken clients to
4318 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4319 unchanged) from the Cygwin maintainer.
4321 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4322 was supported instead of testing for mbx. Effectively a typo.
4324 15. The spa authenticator server code was not checking that the input it
4325 received was valid base64.
4327 16. The debug output line for the "set" modifier in ACLs was not showing the
4328 name of the variable that was being set.
4330 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4332 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4333 lookup on its own. Something else was needed to trigger the lookup. For
4334 example, a match in host_lookup or the need for a host name in a host list.
4335 Now, if $sender_host_name is referenced and the host name has not yet been
4336 looked up, a lookup is performed. If the lookup fails, the variable remains
4337 empty, and $host_lookup_failed is set to "1".
4339 19. Added "eqi" as a case-independent comparison operator.
4341 20. The saslauthd authentication condition could segfault if neither service
4342 nor realm was specified.
4344 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4345 error message that was logged was misleading, and incoming SMTP
4346 connections were dropped. The message is now more accurate, and temporary
4347 errors are given to SMTP connections.
4349 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4350 (except RSET) with a 421 error, until QUIT is received. However, it was
4351 failing to send a response to QUIT.
4353 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4354 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4355 ACL is now run after the helo_try_verify_hosts code.
4357 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4358 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4359 case-independent, so other case variants are also recognized). Apparently
4360 some systems use these upper case variants.
4362 25. If more than two messages were waiting for the same host, and a transport
4363 filter was specified for the transport, Exim sent two messages over the
4364 same TCP/IP connection, and then failed with "socket operation on non-
4365 socket" when it tried to send the third.
4367 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4369 27. The extern definition of crypt16() in expand.c was not being excluded when
4370 the OS had its own crypt16() function.
4372 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4373 as a preferred synonym for return_size_limit, both as an option and as an
4376 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4378 30. Changed the default configuration ACL to relax the local part checking rule
4379 for addresses that are not in any local domains. For these addresses,
4380 slashes and pipe symbols are allowed within local parts, but the sequence
4381 /../ is explicitly forbidden.
4383 31. SPA server authentication was not clearing the challenge buffer before
4386 32. log_message in a "warn" ACL statement was writing to the reject log as
4387 well as to the main log, which contradicts the documentation and doesn't
4388 seem right (because no rejection is happening). So I have stopped it.
4390 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4391 However, I am unable to do any testing of this.
4393 34. Fixed an infelicity in the appendfile transport. When checking directories
4394 for a mailbox, to see if any needed to be created, it was accidentally
4395 using path names with one or more superfluous leading slashes; tracing
4396 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4398 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4399 contents are added to the log line that is written for every discarded
4400 recipient. (Previously a log_message setting was ignored.)
4402 36. The ${quote: operator now quotes the string if it is empty.
4404 37. The install script runs exim in order to find its version number. If for
4405 some reason other than non-existence or emptiness, which it checks, it
4406 could not run './exim', it was installing it with an empty version number,
4407 i.e. as "exim-". This error state is now caught, and the installation is
4410 38. An argument was missing from the function that creates an error message
4411 when Exim fails to connect to the socket for saslauthd authentication.
4412 This could cause Exim to crash, or give a corrupted message.
4414 39. Added isip, isip4, and isip6 to ${if conditions.
4416 40. The ACL variables $acl_xx are now saved with the message, and can be
4417 accessed later in routers, transports, and filters.
4419 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4420 strings in the file are not string-expanded.
4422 42. If a MAIL command specified a SIZE value that was too large to fit into an
4423 int variable, the check against message_size_limit failed. Such values are
4424 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4425 day this will have to be increased, but I don't think I want to be around
4426 when emails are that large.
4433 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4434 iconv() is not standard in FreeBSD.
4436 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4437 IPv6 enabled. The observed symptom was a segmentation fault on return from
4438 the function os_common_find_running_interfaces() in src/os.c.
4440 3. In the check_special_case() function in daemon.c I had used "errno" as an
4441 argument name, which causes warnings on some systems. This was basically a
4442 typo, since it was named "eno" in the comments!
4444 4. The code that waits for the clock to tick (at a resolution of some fraction
4445 of a second) so as to ensure message-id uniqueness was always waiting for
4446 at least one whole tick, when it could have waited for less. [This is
4447 almost certainly not relevant at current processor speeds, where it is
4448 unlikely to ever wait at all. But we try to future-proof.]
4450 5. The function that sleeps for a time interval that includes fractions of a
4451 second contained a race. It did not block SIGALRM between setting the
4452 timer, and suspending (a couple of lines later). If the interval was short
4453 and the sigsuspend() was delayed until after it had expired, the suspension
4454 never ended. On busy systems this could lead to processes getting stuck for
4457 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4458 process, before it forks any delivery processes. The open lookup caching
4459 mechanism meant that the open file or database connection was passed into
4460 the delivery process. The problem was that delivery processes always tidy
4461 up cached lookup data. This could cause a problem for the next delivery
4462 process started by the queue runner, because the external queue runner
4463 process does not know about the closure. So the next delivery process
4464 still has data in the lookup cache. In the case of a file lookup, there was
4465 no problem because closing a file descriptor in a subprocess doesn't affect
4466 the parent. However, if the lookup was caching a connection to a database,
4467 the connection was closed, and the second delivery process was likely to
4468 see errors such as "PGSQL: query failed: server closed the connection
4469 unexpectedly". The problem has been fixed by closing all cached lookups
4470 in a queue runner before running a delivery process.
4472 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4473 seem to have the "const" qualifier which it has on other OS. I've
4476 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4477 *of the same type* (client or server) with the same public name that an
4478 error should be diagnosed.
4480 9. When Exim looked up a host name for an IP address, but failed to find the
4481 original IP address when looking up the host name (a safety check), it
4482 output the message "<ip address> does not match any IP for NULL", which was
4483 confusing, to say the least. The bug was that the host name should have
4484 appeared instead of "NULL".
4486 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4487 user that is built into the binary, and the -C or -D options is used, root
4488 privilege is dropped before the configuration file is read. In addition,
4489 logging is switched to stderr instead of the normal log files. If the
4490 configuration then re-defines the Exim user, the unprivileged environment
4491 is probably not what is expected, so Exim logs a panic warning message (but
4494 However, if deliver_drop_privilege is set, the unprivileged state may well
4495 be exactly what is intended, so the warning has been cut out in that case,
4496 and Exim is allowed to try to write to its normal log files.
4502 1. smtp_return_error_details was not giving details for temporary sender
4503 or receiver verification errors.
4505 2. Diagnose a configuration error if two authenticators have the same public
4508 3. Exim used not to create the message log file for a message until the first
4509 delivery attempt. This could be confusing when incoming messages were held
4510 for policy or load reasons. The message log file is now created at the time
4511 the message is received, and an initial "Received" line is written to it.
4513 4. The automatically generated man page for command line options had a minor
4514 bug that caused no ill effects; however, a more serious problem was that
4515 the procedure for building the man page automatically didn't always
4516 operate. Consequently, release 4.20 contains an out-of-date version. This
4517 shouldn't happen again.
4519 5. When building Exim with embedded Perl support, the script that builds the
4520 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4521 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4523 6. The freeze_tell option was not being used for messages that were frozen on
4524 arrival, either by an ACL or by local_scan().
4526 7. Added the smtp_incomplete_transaction log selector.
4528 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4529 was accepting AUTH without a new EHLO.
4531 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4532 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4535 10. Logging of TCP/IP connections (when configured) now happens in the main
4536 daemon process instead of the child process, so that the TCP/IP connection
4537 count is more accurate (but it can never be perfect).
4539 11. The use of "drop" in a nested ACL was not being handled correctly in the
4540 outer ACL. Now, if condition failure induced by the nested "drop" causes
4541 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4542 or "require"), the connection is dropped.
4544 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4545 that yield "discard" can now be used with an "accept" or a "discard" verb,
4546 but an error is generated for any others (because I can't see a useful way
4547 to define what should happen).
4549 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4550 are now processed in the same way as lines in the Exim configuration file.
4551 In particular, continuation lines are supported.
4553 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4555 15. Added -ti meaning -t -i.
4557 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4558 domains, and warn by logging if others are found.
4560 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4561 returned from GIFCONF when Exim is trying to find the list of interfaces on
4562 a host. The code in os.c has been modified to copy each ifreq to an aligned
4563 structure in all cases.
4565 Also, in some cases, the returned ifreq's were being copied to a 'struct
4566 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4567 means the last couple of bytes of an IPv6 address could be chopped if the
4568 ifreq contained only a normal sockaddr (14 bytes storage).
4570 18. Named domain lists were not supported in the hosts_treat_as_local option.
4571 An entry such as +xxxx was not recognized, and was treated as a literal
4574 19. Ensure that header lines added by a DATA ACL are included in the reject log
4575 if the ACL subsequently rejects the message.
4577 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4578 MD5 (which is deprecated).
4580 21. When testing a filter file using -bf, Exim was writing a message when it
4581 took the sender from a "From " line in the message, but it was not doing so
4582 when it took $return_path from a Return-Path: header line. It now does.
4584 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4585 with a valid header line field name (a series of printing characters
4586 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4588 23. Changed "disc" in the source to "disk" to conform to the documentation and
4589 the book and for uniformity.
4591 24. Ignore Sendmail's -Ooption=value command line item.
4593 25. When execve() failed while trying to run a command in a pipe transport,
4594 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4595 could be confused with a return value of 69 from the command itself. This
4596 has been changed to 127, the value the shell returns if it is asked to run
4597 a non-existent command. The wording for the related log line suggests a
4598 non-existent command as the problem.
4600 26. If received_header_text expands to an empty string, do not add a Received:
4601 header line to the message. (Well, it adds a token one on the spool, but
4602 marks it "old" so that it doesn't get used or transmitted.)
4604 27. Installed eximstats 1.28 (addition of -nt option).
4606 28. There was no check for failure on the call to getsockname() in the daemon
4607 code. This can fail if there is a shortage of resources on the system, with
4608 ENOMEM, for example. A temporary error is now given on failure.
4610 29. Contrary to the C standard, it seems that in some environments, the
4611 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4612 program. Exim now does this explicitly; it affects the formatting of
4613 timestamps using strftime().
4615 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4616 complaints. I've now initialized all the variables, to avoid this.
4618 32. Header lines added by a system filter were not being "seen" during
4619 transport-time rewrites.
4621 33. The info_callback() function passed to OpenSSL is set up with type void
4622 (*)(SSL *, int, int), as described somewhere. However, when calling the
4623 function (actually a macro) that sets it up, the type void(*)() is
4624 expected. I've put in a cast to prevent warnings from picky compilers.
4626 34. If a DNS black list lookup found a CNAME record, but there were no A
4627 records associated with the domain it pointed at, Exim crashed.
4629 35. If a DNS black list lookup returned more than one A record, Exim ignored
4630 all but the first. It now scans all returned addresses if a particular IP
4631 value is being sought. In this situation, the contents of the
4632 $dnslist_value variable are a list of all the addresses, separated by a
4635 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4636 to accept a host name and all its aliases if the forward lookup for any of
4637 them yielded the IP address of the incoming connection. Now it accepts only
4638 those names whose forward lookup yields the correct IP address. Any other
4639 names are discarded. This closes a loophole whereby a rogue DNS
4640 administrator could create reverse DNS records to break through a
4641 wildcarded host restriction in an ACL.
4643 37. If a user filter or a system filter that ran in a subprocess used any of
4644 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4645 the wrong values were passed to the pipe command ($thisaddress had the
4646 value of $0, $0 had the value of $1, etc). This bug was introduced by
4647 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4649 38. Improved the line breaking for long SMTP error messages from ACLs.
4650 Previously, if there was no break point between 40 and 75 characters, Exim
4651 left the rest of the message alone. Two changes have been made: (a) I've
4652 reduced the minimum length to 35 characters; (b) if it can't find a break
4653 point between 35 and 75 characters, it looks ahead and uses the first one
4654 that it finds. This may give the occasional overlong line, but at least the
4655 remaining text gets split now.
4657 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4658 file descriptors might be low, and that setting 1000 would always raise it.
4659 It turns out that in some environments, the limit is already over 1000 and
4660 that lowering it causes trouble. So now Exim takes care not to decrease it.
4662 40. When delivering a message, the value of $return_path is set to $sender_
4663 address at the start of routing (routers may change the value). By an
4664 oversight, this default was not being set up when an address was tested by
4665 -bt or -bv, which affected the outcome if any router or filter referred to
4668 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4669 log only when "message" or "log_message" are set. However, if one of the
4670 conditions was an address verification, or a call to a nested ACL, the
4671 messages generated by the underlying test were being passed through. This
4672 no longer happens. The underlying message is available in $acl_verify_
4673 message for both "message" and "log_message" expansions, so it can be
4674 passed through if needed.
4676 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4677 new expansion $bh_ to give the encoded byte string without charset
4678 translation. Translation happens only if iconv() is available; HAVE_ICONV
4679 indicates this at build time. HEADERS_CHARSET gives the charset to
4680 translate to; headers_charset can change it in the configuration, and
4681 "headers charset" can change it in an individual filter file.
4683 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4684 that creates RFC 2047 encoded "words" labels them as that charset instead
4685 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4686 expansion operator; (ii) when Exim creates a From: line for a local
4687 message; (iii) when a header line is rewritten to include a "phrase" part.
4689 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4690 buggy, causing it to skip the first lines of messages whose message ID
4691 ended in 'D'. This would not have bitten before Exim release 4.14, because
4692 message IDs were unlikely to end in 'D' before then. The effect was to have
4693 incorrect size information for certain domains.
4695 45. #include "config.h" was missing at the start of the crypt16.c module. This
4696 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4699 46. If there was a timeout during a "random" callout check, Exim treated it as
4700 a failure of the random address, and carried on sending RSET and the real
4701 address. If the delay was just some slowness somewhere, the response to the
4702 original RCPT would be taken as a response to RSET and so on, causing
4703 mayhem of various kinds.
4705 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4706 when I implemented it. It didn't allow for the fact that some option values
4707 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4708 the right test for positive values.
4710 48. Domain names in DNS records are case-independent. Exim always looks them up
4711 in lower case. Some resolvers return domain names in exactly the case they
4712 appear in the zone file, that is, they may contain uppercase letters. Not
4713 all resolvers do this - some return always lower case. Exim was treating a
4714 change of case by a resolver as a change of domain, similar to a widening
4715 of a domain abbreviation. This triggered its re-routing code and so it was
4716 trying to route what was effectively the same domain again. This normally
4717 caused routing to fail (because the router wouldn't handle the domain
4718 twice). Now Exim checks for this case specially, and just changes the
4719 casing of the domain that it ultimately uses when it transmits the message
4722 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4725 50. If a filter generated a file delivery with a non-absolute name (possible if
4726 no home directory exists for the router), the forbid_file option was not
4729 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4730 the existing equality matching.
4732 52. Exim was using ints instead of ino_t variables in some places where it was
4733 dealing with inode numbers.
4735 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4736 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4737 called TMPDIR, and if it finds it is different, it changes its value.
4739 54. The smtp_printf() function is now made available to local_scan() so
4740 additional output lines can be written before returning. There is also an
4741 smtp_fflush() function to enable the detection of a dropped connection.
4742 The variables smtp_input and smtp_batched_input are exported to
4745 55. Changed the default runtime configuration: the message "Unknown user"
4746 has been removed from the ACL, and instead placed on the localuser router,
4747 using the cannot_route_message feature. This means that any verification
4748 failures that generate their own messages won't get overridden. Similarly,
4749 the "Unrouteable address" message that was in the ACL for unverifiable
4750 relay addresses has also been removed.
4752 56. Added hosts_avoid_esmtp to the smtp transport.
4754 57. The exicyclog script was not checking for the esoteric option
4755 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4756 will work only if exicyclog is run under the appropriate euid.
4758 58. Following a discussion on the list, the rules by which Exim recognises line
4759 endings on incoming messages have been changed. The -dropcr and drop_cr
4760 options are now no-ops, retained only for backwards compatibility. The
4761 following line terminators are recognized: LF CRLF CR. However, special
4762 processing applies to CR:
4764 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4765 nor a local message in the state where . is a terminator.
4767 (ii) If a bare CR is encountered in a header line, an extra space is added
4768 after the line terminator so as not to end the header. The reasoning
4769 behind this is that bare CRs in header lines are most likely either
4770 to be mistakes, or people trying to play silly games.
4772 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4773 was being incorrectly given as 18 bytes larger than it should have been.
4774 This is a VOB (very old bug).
4776 60. This may never have affected anything current, but just in case it has:
4777 When the local host is found other than at the start of a list of hosts,
4778 the local host, those with the same MX, and any that follow, are discarded.
4779 When the list in question was part of a longer list of hosts, the following
4780 hosts (not currently being processed) were also being discarded. This no
4781 longer happens. I'm not sure if this situation could ever has previously
4784 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4787 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4788 header. This is something that is recommended in a new Internet Draft, and
4789 is something that is documented as being done by Sendmail. There are two
4790 possible values. For messages generated by the autoreply transport, Exim
4793 Auto-Submitted: auto-replied
4795 whereas for all other generated messages (e.g. bounces) it adds
4797 Auto-Submitted: auto-generated
4799 63. The "personal" condition in filters now includes a test for the
4800 Auto-Submitted: header. If it contains the string "auto-" the message it
4801 not considered personal.
4803 64. Added rcpt_include_affixes as a generic transport option.
4805 65. Added queue_only_override (default true).
4807 66. Added the syslog_duplication option.
4809 67. If what should have been the first header line of a message consisted of
4810 a space followed by a colon, Exim was mis-interpreting it as a header line.
4811 It isn't of course - it is syntactically invalid and should therefore be
4812 treated as the start of the message body. The misbehaviour could have
4813 caused a number of strange effects, including loss of data in subsequent
4814 header lines, and spool format errors.
4816 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4817 client host had authenticated. This control can now be exercised by an ACL
4818 for more flexibility.
4820 69. By default, callouts do not happen when testing with -bh. There is now a
4821 variant, -bhc, which does actually run the callout code, including
4822 consulting and updating the callout cache.
4824 70. Added support for saslauthd authentication, courtesy of Alexander
4827 71. If statvfs() failed on the spool or log directories while checking their
4828 size for availability, Exim confusingly gave the error "space shortage".
4829 Furthermore, in debugging mode it crashed with a floating point exception.
4830 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4831 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4832 serious problem, Exim now writes to the main and panic logs when this
4833 happens, with details of the failure. It then refuses to accept the
4834 incoming message, giving the message "spool directory problem" or "log
4835 directory problem" with a 421 code for SMTP messages.
4837 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4838 0, 1, and 2 exist, because some OS complain for execs without them (see
4839 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4840 descriptors. However, the code omitted to check that the open succeeded,
4841 causing mysterious errors if for some reason the permissions on /dev/null
4842 got screwed. Now Exim writes a message to the main and panic logs, and
4843 bombs out if it can't open /dev/null.
4845 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4846 interact so that it is all more flexible. It is supposed to remain
4847 backwards compatible. Also added extra_local_interfaces.
4849 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4850 to bomb out with an assertion failure - to the client this appears as a
4851 connection drop. This problem occurs in the part of the code that was taken
4852 from the Samba project. Fortunately, the assertion is in a very simple
4853 function, so I have fixed this by reproducing the function inline in the
4854 one place where it is called, and arranging for authentication to fail
4855 instead of killing the process with assert().
4857 75. The SPA client code was not working when the server requested OEM rather
4858 than Unicode encoding.
4860 76. Added code to make require_files with a specific uid setting more usable in
4861 the case where statting the file as root fails - usually a non-root-mounted
4862 NFS file system. When this happens and the failure is EACCES, Exim now
4863 forks a subprocess and does the per-uid checking as the relevant uid.
4865 77. Added process_log_path.
4867 78. If log_file_path was not explicitly set, a setting of check_log_space or
4868 check_log_inodes was ignored.
4870 79. If a space check for the spool or log partitions fails, the incident is now
4871 logged. Of course, in the latter case the data may get lost...
4873 80. Added the %p formatting code to string_format() so that it can be used to
4874 print addresses in debug_print(). Adjusted all the address printing in the
4875 debugging in store.c to use %p rather than %d.
4877 81. There was a concern that a line of code in smtp_in.c could overflow a
4878 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4879 initially expressed, the concern was not well-founded, because trailing
4880 spaces are removed early. However, if the trailing spaces were followed by
4881 a NULL, they did not get removed, so the overflow was possible. Two fixes
4884 (a) I re-wrote the offending code in a cleaner fashion.
4885 (b) If an incoming SMTP command contains a NULL character, it is rejected
4888 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4889 runs initgroups(), so that if the Exim user is in any additional groups,
4890 they will be used during message reception.
4896 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob